Gondwana Research 79 (2020) 283e300

Contents lists available at ScienceDirect

Gondwana Research
journal homepage: www.elsevier.com/locate/gr

Early central American forearc follows the subduction initiation rule

Scott A. Whattam a, *, Camilo Montes b, Robert J. Stern c
a
Department of Geosciences, King Fahd University of Petroleum and Minerals, Dhahran, 31261, Saudi Arabia
b
Department of Physics and Geology, Universidad del Norte, Barranquilla, Colombia
c
Department of Geosciences, University of Texas at Dallas, Richardson, TX, USA

a r t i c l e i n f o a b s t r a c t

Article history: The “subduction initiation rule” (SIR) (Whattam and Stern, 2011) advocates that proto-arc and forearc
Received 12 October 2018 complexes preserved in ophiolites and forearcs follow a predictable chemotemporal and/or chemo-
Received in revised form stratigraphic vertical progression. This chemotemporal evolution is defined by a progression from bot-
11 September 2019
tom to top, from less to more depleted and slab-metasomatized sources. This progression has been
Accepted 2 October 2019
Available online 9 November 2019
recently documented for other igneous suites associated with subduction initiation. The Sona-Azuero
forearc complex of southern Panama represents the earliest magmatic arc activity at the Central
American Volcanic Arc system. Comparison of new and existing geochemical data for the circa 82-40 Ma
Keywords:
Central American forearc
Sona-Azuero Proto-Arc/Arc, its underlying 89-85 Ma “oceanic plateau” of SW Panama and the 72-69 Ma
Subduction initiation Golfito Proto-Arc of southern Costa Rica with the 70-39 Ma Chagres-Bayano Arc of eastern Panama
Subduction initiation rule exhibits a chemotemporal progression as described above and which follows the SIR. Sona-Azuero lavas
Forearc are predominantly MORB-like, whereas those of the younger Chagres-Bayano complex are mostly VAB-
Volcanic arc like; lavas of the Golfito Proto-Arc typically show characteristics intermediate to that of the Sona-Azuero
Ophiolite and Chagres-Bayano proto-arc/arc complexes. On the basis of isotope evidence as shown in other studies,
lava types of all three complexes are clearly derived from a source contaminated by the Caribbean Large
Igneous Province plume; we term these “plume-contaminated” forearc basalts and volcanic arc basalts,
respectively. Apart from a plume-induced subduction initiation origin for the Panamanian forearc, these
insights suggest otherwise similar petrogenetic origins and tectonic setting to lavas comprising earliest-
formed forearc crust, and most ophiolites, which follow the SIR.
© 2019 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.

1. Introduction (VAB) or basaltic-andesite lavas and intrusives characterized by

Ophiolites are ancient, on-land vestiges of oceanic lithosphere
ascribed to seafloor spreading to form a nascent proto-forearc in
the earliest stages of subduction initiation. As shown by Whattam
and Stern (2011), Tethyan-type supra-subuduction zone ophio-
lites (Pearce et al., 1984; Pearce, 2003) of the Eastern-
Mediterranean-Persian Gulf region (Robertson, 2004) exhibit
similar chemostratigraphic/chemotemporal progressions (changes
in lava chemistry with time) as that of forearc lithosphere of classic
records of magmatism during subduction initiation, such as the Izu-
Bonin-Mariana intra-oceanic arc system (Reagan et al., 2010;
Ishizuka et al., 2011). Tethyan-type ophiolites for example, in Oman
(Semail), Greece (Pindos) and Albania (Mirdita) formed by initial
eruption of MOR-like tholeiitic basalts (i.e., forearc basalts, FAB,
Reagan et al., 2010) followed by calc-alkaline volcanic arc basalt
* Corresponding author.
E-mail addresses: sawhatta@gmail.com, scott.whattam@kfupm.ed.sa
(S.A. Whattam).
large ion lithophile-enrichment and high-field strength element-
depletion, which were sometimes followed by latest-stage bonin-
ites (e.g., Alabaster et al., 1982; Saccani and Photiades, 2004; Dilek
et al., 2008). This exact sequence is not universal but a common
theme is emerging from global examples. Variations in subduction
initiation progression is revealed by the Troodos ophiolite which
lacks MORB-like tholeiites and instead shows a progression from
earlier-formed basalt-andesite-dacite-rhyolite series backarc basin-
like basalts to boninites (Pearce and Robinson, 2010). At the Izu-
Bonin-Mariana forearc, generation of earliest MORB-like FAB dur-
ing initial subduction initiation was followed by eruption of
boninites (Reagan et al., 2010, 2017); from this time forward
igneous activity retreated to the site of the long-term magmatic
front. A similar chemotemporal progression has recently been
documented for the nascent stages of Greater Antilles Arc system
construction (Torro
et al., 2017). This chemotemporal progression
reflects complementary magmatic and mantle responses (MORB /
VAB affinities and fertile lherzolite / depleted harzburgite,
respectively) during subduction initiation and encapsulates the

https://doi.org/10.1016/j.gr.2019.10.002
1342-937X/© 2019 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.
 284 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300

(2010); b, is to indicate that this refers to a restricted segment of the Azuero Arc (the Río Quema Formation, Corral et al., 2011); c, Golfito also consists of basaltic trachyandesite and andesite according to Buchs et al. (2010) and
Age constraints for the Golfito, Sona-Azuero and Chagres-Bayano complexes. Abbreviations: A, arc; CR, Costa Rica; PA, proto-arc; PAN, Panama; OP, oceanic plateau. Superscripts: a, Azuero Arc includes the proto-arc of Buchs et al.

references therein; a TAS plot (not shown) does suggest this, however a plot of Nb/Y vs. Zr/Ti (not shown) suggests that designation of alkaline affinities is the result of alteration (addition of alkalis); d, similarly, the Nb/Y vs. Zr/Ti plot
suggests no alkaline affinities; e, Wegner et al. (2011) suggested alteration could account for the high K2O contents of some rock types which is confirmed by a Nb/Y vs. Zr/Ti plot; f and g, interpreted by Hauff et al. (2000) and Buchs

Marco et al. (1995);Hauff et al. (2000);

Dengo (1962); Schmidt-Effing (1979);

Giudice and Recchi (1969); Metti and

Recchi (1976); Kolarsky et al. (1995);

(2002, 2006); Lissinna (2005); Buchs

(2009); Corral et al., (2011); Wegner

Obando (1986); Di Marco (1994); Di

Hoernle et al., (2002); Lissinna et al.

(2005, 2009); Wegner et al., (2011);

et al., (2011); Montes et al. (2012a)

€ rner et al.
€rner et al.,

Montes et al. (2012a, 2012b)

Maury et al. (1995); Wo
et al. (2009, 2010); Wo
Buchs et al. (2010)
oceanic plateau

oceanic plateau
and proto-arc/

proto-arc/arc
arc

arc
radiometric (40Ar/39Ar,

radiometric (40Ar/39Ar,
Biostratigraphic fossil

Biostratigraphic fossil
biostratigraphic fossil

LA-ICP-MS UePb on

LA-ICP-MS UePb on
(foraminifera) ID;
(radiolarite) ID

zircons)

zircons)

Fig. 1. (a) Present tectonic configuration of the Circum-Caribbean region. (b) Magni-
ID

fication and detail of the boxed region in Fig. 1a showing the distribution of Late
Cretaceous to middle Eocene (~75-39 Ma) arc magmatism in Golfito, easternmost
Late Cretaceous to Early

Costa Rica, Sona-Azuero, southern Panama and Chagres-Bayano, eastern Panama.

middle Maastrichian

Santonian (~89e85)

Maastrichian 82-40

Abbreviations in: (a) CR, Costa Rica,; GAA, Greater Antilles Arc; LAA, Lesser Antilles Arc;
late Campanian to

PAN, Panama; PR, Puerto Rico. (b) CB, Chagres-Bayano; CI, Coiba Island; Cord de Pan,
Coniacian-Early

Cordillera of Panama; GF, Golfito; PC, Panama City; PCFZ, Panama Canal Fracture Zone.
(~72-69 Ma)

70e39

‘subduction initiation rule’ (SIR) of Whattam and Stern (2011). The

SIR stipulates that the composition of subduction initiation lavas
change from less to more high-field strength depleted and from less
sheet flows, pillow basalts,

massive and pillowed lava

overlying and intercalated

to more slab-metasomatized with time, i.e., lavas change from

massive to thin pillowed

flow-dyke complex with

rare gabbroic intrusives

associated sub-volcanic
lavas flows, dykes, lava
domes, large intrusive

volcaniclastic breccias

MORB-like to VAB-like over the course of subduction initiation, and

plutonic rocks with

this rule links forearcs formed during subduction initiation and

ophiolites. This chemotemporal progression evolution reflects for-
mation of proto-forearc/forearc lithosphere as a result of mantle
complexes
lava flows

melting that is increasingly influenced by enrichments from the

sinking slab during subduction initiation. Mid-oceanic ridge-like
FAB lavas are the consequence of decompression melting of up-
welling asthenosphere and mark the initial magmatic manifesta-
enriched OIB-like varieties)
subalkaline basalt, (minor)

subalkaline basalt, (minor)

tholeiitic basalt to rhyolite
subalkaline basalt, basaltic

tion of subduction initiation. Contribution of fluids from

gabbro basalt to rhyolite
equilvalents (including

dehydrating oceanic crust and sediments on the sinking slab is

minor during the early stages of subduction initiation, but contin-
and their plutonic

and their plutonic

trachyandesite,

uous melting results in a depleted, harzburgitic residue. This res-

trachyandesite
et al. (2010) as oceanic plateau and proto-arc/arc, respectively.

equilvalents

idue is increasingly metasomatized by fluids from the sinking slab

and later partial melting of this residue yields ‘typical’
gabbro

suprasubduction-zone-like lavas with high-field strength element

depletions and large ion element enrichments in the latter stages of
subduction initiation. Recognizing the SIR has profound implica-
~1.7

tions for the development of robust tectonic plate reconstructions

~3
?

and interpreting and understanding forearc formation during

subduction initiation. Identification of the SIR in ophiolite com-
submarine

submarine

submarine

plexes implies formation similar to that of forearc lithosphere via

seafloor spreading during subduction initiation (Stern and Bloomer,
1992; Stern et al., 2012).
Table 1 Whattam, 2019.

The purpose of this paper is not to provide a regional treatment

Azuero Proto-Arc &

Arc (E. Panama)

of the tectonic evolution of or new stratigraphic constraints on the

Azuero Plateau
southern Costa
(i.e.,Proto-Arc,

Azuero Complex
Golfito Complex

Chagres-Bayano
(S. Panama)

early Central American forearc but rather, to examine the early

Central American forearc as an extension to cases for which the SIR
Rica)

has been tested so far. We do this by focusing on Late Cretaceous

Arc
Table 1

and Early Paleogene igneous rocks of Central America and compare

their chemotemporal (changes in chemistry with time) progression
 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300
Fig. 2. Geologic map of the Azuero peninsula showing the location of our samples. Map is modified from Direccio
Corral et al. (2011). See Fig. 1 for broader location of the Azuero peninsula. Abbreviations: CB, Chagres-Bayano; G, Group: GI, Group I; GII, Group II; GIII, Group III; Nic, Nicaragua; SA,
Sona-AzueroSee text for definitions of these groups.
with the chemotemporal progression of other complexes associ-
ated with subduction initiation (forearcs and ophiolites).
Crucial to our chemotemporal arguments is recognition that the
Sona-Azuero and Golfito proto-arc/arc segments in central Panama
and SW Costa Rica, respectively, began to form at least 5 M.y. prior
to establishment of the Chagres-Bayano Arc in central Panama (see
Section 2 for Regional setting and the Supplementary Document for
description of these complexes). We document the availability of
absolute and relative ages which provides the basis of this recog-
nition in the Supplementary Document (Table S1, see also Table 1).
2. Regional setting
As a result of detailed multidisciplinary studies of Late Creta-
ceous ‘oceanic igneous terranes’ (terminology of Hauff et al., 1997)
and arc-related complexes exposed in the forearc of the Central
American Volcanic Arc (Fig. 1a) in Costa Rica (e.g., Buchs et al., 2010)
and Panama (Fig. 1b) (e.g., Hauff et al., 2000a; Tomascak et al.,
2000; Geldmacher et al., 2001, 2008; Lissinna, 2005; Gazel et al.,
2009, 2015; Wo €rner et al., 2009; Buchs et al., 2009, 2010; Wegner
et al., 2011; Montes et al., 2012a, 2012b; Whattam and Stern,
2015a, 2015b) we can re-evaluate the chemotemporal evolution
of these oldest Central American Volcanic forearc complexes in the
context of subduction initiation processes. Most researchers agree
that the bulk of the Caribbean Plate is composed of the Caribbean
Large Igneous Province (CLIP) (Fig. 1a), an expansive oceanic
285

n General de Recursos Minerales (1991), Buchs et al. (2010) and
“plateau” that formed over the Galapagos hotspot prior to its
northeasterly drift and insertion between North and South America
(see Kerr et al., 2000) prior to subduction initiation. Relatedly, the
CLIP is interpreted by most as underlying younger arc segments not
only in Costa Rica and Panama but also along NW South America in
western Ecuador and Colombia all the way to the Leeward Antilles
in Curaçao and Aruba (see references in Whattam and Stern,
2015a). The ‘standard’ model holds that volcanic arc construction
occurred upon the trailing edge of the CLIP in present-day Costa
Rica and Panama (Pindell and Kennan, 2009; Wo €rner et al., 2009;
Buchs et al., 2010; Wegner et al., 2011) and elsewhere in the
Caribbean realm along the southern margin of the Caribbean Plate
in NW South America and the Leeward Antilles. In contrast, on the
basis of collated geochemical, isotopic and radioisotopic age data-
sets of these ‘plume-and-arc-related’ complexes in Central America
and NW South America which exhibit compositional and temporal
overlap between Late Cretaceous hotspot-derived CLIP magmatism
and the magmatic effects of subduction initiation, Whattam and
Stern (2015a) suggested arc construction in a single, rapidly
evolving hybrid plume- and subduction-related environment. By
‘plume- and subduction-related’ we mean proto-forearc/forearc
complexes that were generated synchronously or nearly so as a
result of interaction between the plumehead and the (soon to be)
subduction-modified lithosphere that it interacted with; as a result,
the complexes which constitute this study were generated from a
mantle source that was plume-dominated at first and evolved with
286 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300

Table 2
Whole rock ICP-AES analyzes of Groups I, II, and III Azuero lavas from this study. Abbreviation: bas-and, basaltic-andesite.

Sample No. 300070 300071 300075 300076 300077 300078 300079 300081 300082 300082 300084 300085 Sample No. 300086 300088
Lithology basalt basalt basalt basalt basalt basalt basalt basalt basalt (Rep) basalt basalt Lithology basalt basalt
Groupa Group I Group I Group I Group I Group I Group I Group I Group I Group I basalt Group I Group I Groupa Group I Group I
Group I

major oxides (wt. %) major oxides (wt. %)

SiO2 49.23 48.92 47.29 47.38 49.70 47.00 46.94 46.87 46.61 46.4 45.11 48.00 SiO2 47.21 47.93
TiO2 1.22 1.20 1.22 1.23 1.24 1.22 1.26 1.09 1.22 1.23 0.88 0.95 TiO2 1.22 1.25
Al2O3 13.55 13.42 13.92 13.28 13.67 13.70 14.06 14.03 14.02 13.98 13.39 14.10 Al2O3 14.03 16.75
Fe2O3 11.76 11.74 12.02 11.58 10.80 11.92 12.12 11.16 11.46 11.53 10.29 10.50 Fe2O3 12.12 11.04
MnO 0.18 0.18 0.21 0.23 0.18 0.22 0.23 0.26 0.31 0.30 0.18 0.18 MnO 0.20 0.20
MgO 7.38 8.38 7.26 8.05 7.17 8.13 7.01 8.32 7.84 7.99 8.59 8.40 MgO 8.10 6.40
CaO 10.33 9.91 12.76 12.40 9.72 12.20 12.51 11.96 11.71 11.71 9.83 9.82 CaO 11.78 9.48
Na2O 3.60 3.29 2.09 2.00 2.98 2.23 2.53 2.30 2.16 2.17 2.31 3.56 Na2O 2.39 3.29
K2O 0.16 0.18 0.07 0.07 1.74 0.10 0.09 0.07 0.07 0.08 0.57 0.28 K2O 0.14 0.67
P2O5 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.09 0.09 0.09 0.07 0.07 P2O5 0.10 0.11
LOI 2.20 2.40 2.80 3.40 2.40 2.90 2.90 3.60 4.20 4.20 8.50 3.80 LOI 2.40 2.60
total 99.71 99.72 99.74 99.72 99.70 99.72 99.75 99.75 99.69 99.77 99.72 99.66 total 99.69 99.72
Mg# 56 59 55 58 57 58 54 60 58 58 62 61 Mg# 57 54
trace elements (ppm) trace elements (ppm)
Sc 47 48 48 50 48 51 50 48 50 50 46 47 Sc 47 44
V 328 316 326 331 308 336 324 289 310 314 275 325 V 343 308
Ni 68.1 46.4 76.2 69.6 76.0 69.4 85.5 65.8 66.4 NA 51.5 64.3 Ni 48.0 46.4
Cu 140.4 138.1 143.3 142.8 144.4 140.1 145.8 128.3 138.0 NA 109.7 88.0 Cu 138.3 182.7
Zn 56 67 54 61 60 49 57 50 57 NA 49 36 Zn 56 65
Rb 2.4 2.4 0.7 0.6 18.5 1.1 1.0 0.7 0.7 0.6 1.1 3.0 Rb 1.7 7.3
Sr 233.9 204.8 162.7 118.1 209.4 155.4 108.1 142.2 156.7 158.8 134.9 193.3 Sr 143.2 296.3
Y 21.9 21.6 22.7 22.6 20.6 22.4 22.0 18.7 19.9 20.8 16.0 19.0 Y 23.0 23.8
Zr 59.2 58.7 84.1 61.8 57.0 58.2 56.7 66.1 52.4 63.1 41.8 43.1 Zr 59.1 68.4
Nb 3.3 3.9 3.6 3.7 3.3 3.2 3.0 2.9 3.3 3.1 3.2 2.4 Nb 3.6 3.3
Cs <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 Cs 0.1 0.3
Ba 37 50 14 15 204 17 17 60 27 31 18 151 Ba 60 140
La 3.2 3.6 3.2 3.3 2.8 3.1 3.0 2.6 2.8 2.9 2.3 2.5 La 3.2 3.3
Ce 8.4 8.7 8.7 8.7 7.7 7.7 8.0 6.8 7.6 7.4 5.6 7.2 Ce 9.1 8.7
Pr 1.37 1.4 1.42 1.41 1.32 1.35 1.34 1.14 1.22 1.24 0.97 1.04 Pr 1.42 1.39
Nd 7.5 7.3 7.1 7.4 6.5 7.5 7.0 5.9 6.4 6.6 5.1 6.4 Nd 7.1 7.4
Sm 2.4 2.35 2.51 2.47 2.19 2.35 2.23 1.99 2.17 2.11 1.62 1.93 Sm 2.14 2.41
Eu 0.90 0.91 0.97 0.92 0.90 0.92 0.89 0.79 0.89 0.89 0.63 0.72 Eu 0.99 0.92
Gd 3.18 3.25 3.39 3.40 2.97 3.24 3.17 2.75 3.03 2.96 2.27 2.44 Gd 3.14 3.37
Tb 0.63 0.63 0.64 0.64 0.59 0.62 0.60 0.54 0.57 0.58 0.44 0.49 Tb 0.61 0.65
Dy 3.78 3.80 3.76 3.86 3.77 3.90 3.92 3.24 3.49 3.54 2.71 3.22 Dy 3.66 4.22
Ho 0.80 0.82 0.82 0.83 0.79 0.82 0.77 0.69 0.74 0.73 0.59 0.67 Ho 0.85 0.89
Er 2.34 2.37 2.44 2.41 2.30 2.40 2.35 2.04 2.06 2.15 1.69 2.00 Er 2.62 2.68
Tm 0.36 0.36 0.36 0.35 0.35 0.37 0.36 0.31 0.33 0.31 0.26 0.29 Tm 0.34 0.39
Yb 2.28 2.36 2.30 2.33 2.35 2.30 2.33 1.97 2.15 2.25 1.84 1.79 Yb 2.29 2.55
Lu 0.35 0.36 0.36 0.35 0.34 0.33 0.34 0.30 0.33 0.33 0.26 0.27 Lu 0.34 0.39
Hf 1.9 2 2.3 2.0 2.0 1.8 1.8 2.0 1.7 1.5 1.2 1.3 Hf 1.8 1.9
Ta 0.2 0.2 0.2 0.3 0.2 0.2 0.2 0.2 0.3 0.1 0.1 0.2 Ta 0.2 0.2
Pb 0.6 0.2 0.6 0.2 0.2 0.3 0.3 0.2 <0.1 NA 0.1 <0.1 Pb 0.3 0.2
Th 0.3 <0.2 0.3 0.2 0.4 <0.2 0.2 <0.2 <0.2 0.2 0.2 <0.2 Th <0.2 0.4
U 0.1 0.1 0.1 0.2 0.1 0.1 0.1 <0.1 0.1 0.1 <0.1 <0.1 U <0.1 0.1

Major element detection limits are about 0.001e0.4%. Other notes: Samples Group numbers in italics are samples with total oxide concentrations <96 wt % and which
were thus excluded from our chemical plots (see Section 4.3.2.).
NA ¼ Not analyzed. Rep ¼ Replicate (analysis). STD ¼ Standard.
Table 2 Whattam, 2019.
a
Group refers to classification on the basis of petrographic characteristics (see Supplementary Document) and trace element chemistry (see Section 5.1 and Supple-
mentary Document).

time to be modified by subduction. According to this model, arc
formation occurred subsequent to subduction initiation which was
catalyzed by plume emplacement (plume-induced subduction
initiation). A plume-induced subduction initiation mechanism has
recently been shown to be valid on the basis of numerical ther-
momechanical modelling (Gerya et al., 2015).
Although the Farallon Plate has been subducting beneath the
western Americas since at least 200 Ma, subduction initiation at
the Central American Arc was independent of Farallon Plate
subduction. Central America subduction initiation began between
90 and 75 Ma along the western periphery of the Caribbean Large
Igneous Province (CLIP), which was emplaced within the middle
of the Farallon Plate beginning about 92 Ma (Sinton et al., 1997;
Kerr et al., 2003 and references therein). Some previous studies of
the Central American forearc igneous complexes discussed in this
study (see Supplementary Document] suggest that subduction
initiation began ~75-71 Ma with the magmatic arc growing on
pre-existing oceanic plateau (CLIP) basement (Buchs et al., (2010);
Wegner et al., (2011))_. In contrast, Lissinna et al. (2006) suggest
that the infant Panamanian island arc on the western side of the
CLIP dates back to 89 Ma. The studies of Lissinna et al. (2006) and
Wegner et al. (2011) did not posit a cause for subduction initia-
tion but Buchs et al. (2010) suggested this was tectonically
induced, caused by compression of the thickened Caribbean Plate
during westward migration of the Americas and collision with
South America. As explained above, Whattam and Stern (2015a)
in contrast suggested that subduction initiation was the result
of plume emplacement and discuss why an induced subduction
 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300 287

300089 300091 300072 300073 300074 300083 300066 300067 300068 Sample No. 300069 300087 300090 STD STD STD STD
basalt basalt basalt basalt basalt basalt andesite basalt basalt Lithology bas-and basalt basalt SO- SO-18 SO-18 SO-18
Group I Group I Group II Group II Group II Group II Group III Group III Group III Groupa Group III Group III Group III 18 Rep 1 Rep 2 Rep 3

major oxides (wt. %) major oxides (wt. %)

48.78 47.04 44.77 43.97 46.95 45.00 56.28 48.11 47.06 SiO2 54.30 47.40 50.34 58.29 58.22 58.13 58.11
1.76 1.26 4.55 4.69 2.76 3.64 1.14 0.57 0.51 TiO2 0.19 1.74 0.44 0.7 0.7 0.69 0.69
13.72 13.86 11.78 11.82 13.34 13.09 15.37 20.30 18.78 Al2O3 11.39 14.29 18.34 13.98 13.98 14.11 14.13
13.18 11.74 16.96 16.68 14.18 13.37 9.05 8.79 10.11 Fe2O3 11.49 12.70 10.06 7.59 7.59 7.58 7.58
0.21 0.21 0.23 0.21 0.22 0.17 0.17 0.15 0.19 MnO 0.19 0.22 0.20 0.4 0.4 0.39 0.39
6.23 7.55 5.52 5.46 6.51 5.62 3.46 4.91 7.24 MgO 7.40 5.85 5.28 3.34 3.35 3.35 3.33
10.80 12.64 8.34 11.06 9.82 9.46 6.37 10.80 12.99 CaO 7.84 8.85 9.80 6.31 6.35 6.36 6.38
2.74 1.80 3.46 2.15 3.25 3.86 3.94 3.23 1.63 Na2O 2.20 3.49 2.25 3.71 3.7 3.7 3.71
0.17 0.04 0.46 0.51 0.33 0.92 0.99 0.47 0.10 K2O 0.29 1.14 0.32 2.15 2.16 2.15 2.16
0.18 0.10 0.44 0.45 0.24 0.43 0.31 0.07 0.05 P2O5 0.07 0.17 0.10 0.83 0.84 0.82 0.81
2.00 3.50 3.20 2.70 2.10 4.20 2.80 2.40 1.10 LOI 4.30 3.80 2.60 1.9 1.9 1.9 1.9
99.77 99.74 99.71 99.70 99.70 99.76 99.88 99.80 99.76 total 99.66 99.65 99.73 99.76 99.76 99.74 99.74
48 56 39 39 48 46 43 53 59 Mg# 56 48 51 47 47 47 47
trace elements (ppm) trace elements (ppm)
53 48 38 38 39 28 31 33 43 Sc 59 44 41 26 26 25 25
351 309 462 463 426 336 193 241 296 V 338 336 267 192 191 205 207
17.3 73.6 34.8 34.6 51.1 56.0 2.3 16.3 9.9 Ni 20.3 19.2 9.1 46 45 46 43
100.1 135.4 317.7 323.7 140.9 25.1 42.4 151.5 221.6 Cu 71.7 107.8 118.1 NA NA NA NA
64 52 139 107 88 75 57 40 17 Zn 128 75 35 NA NA NA NA
1.2 0.4 5.0 7.1 4.5 16.5 11.6 7.2 2.1 Rb 3.2 14.5 8.2 26.4 26.4 27.8 28
146.0 172.0 72.7 155.9 218.3 473.0 229.5 373.7 360.7 Sr 164.8 336.9 764.3 377.8 380.7 399 413.7
37.4 21.4 51.2 53.4 32.0 29.5 31.7 11.6 10.7 Y 8.8 32.9 13.5 29 29.7 31.3 31.4
105.0 63.3 272.2 281.5 141.1 237.6 101.9 25.5 11.0 Zr 15.0 100.2 25.6 261.1 264 278.9 282.7
2.2 3.8 20.3 20.9 12.0 32.7 3.3 1.2 0.8 Nb 1.8 2.5 0.5 18.9 19.1 19.4 20.5
<0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 0.3 0.1 Cs <0.1 <0.1 0.1 6.5 6.7 6.7 7
71 12 51 59 92 140 397 159 72 Ba 668 1100 219 476 484 500 513
4.2 3.3 17.4 17.7 9.8 23.4 10.9 2.6 2.3 La 3.2 5.0 3.4 11.2 11.4 12.1 12.2
12.0 8.2 45.7 46.8 24.5 55 24.0 5.6 4.6 Ce 4.9 13.8 6.9 24.7 25.2 26.9 27.5
2.13 1.36 6.87 7.31 3.77 7.27 3.51 0.89 0.73 Pr 0.90 2.32 1.16 3.15 3.16 3.27 3.31
12.6 7.1 34 34 18.1 33 16.4 4.2 3.4 Nd 4.2 12.2 5.9 13.1 12.7 13.6 13.4
4.00 2.28 9.23 9.74 5.33 7.21 4.28 1.29 1.28 Sm 1.05 3.65 1.62 2.73 2.71 2.84 2.84
1.41 0.88 2.83 2.91 1.85 2.39 1.37 0.57 0.58 Eu 0.35 1.42 0.57 0.81 0.84 0.84 0.84
5.44 3.15 10.75 10.94 6.21 7.47 5.23 1.74 1.69 Gd 1.24 4.87 2.03 2.77 2.8 2.95 2.94
1.04 0.62 1.76 1.83 1.06 1.11 0.89 0.33 0.32 Tb 0.22 0.90 0.34 0.48 0.47 0.5 0.49
6.42 3.75 10.06 10.05 6.07 6.00 5.63 2.02 1.91 Dy 1.26 5.70 2.13 2.74 2.73 2.86 2.88
1.38 0.78 1.84 1.97 1.20 1.05 1.16 0.44 0.42 Ho 0.28 1.18 0.48 0.58 0.58 0.6 0.61
4.11 2.29 5.02 5.22 3.24 2.80 3.43 1.32 1.19 Er 0.85 3.48 1.44 1.71 1.7 1.71 1.79
0.62 0.34 0.68 0.71 0.47 0.37 0.51 0.21 0.20 Tm 0.15 0.52 0.23 0.27 0.26 0.27 0.27
4.03 2.22 4.36 4.65 3.07 2.39 3.51 1.31 1.29 Yb 1.05 3.40 1.54 1.63 1.69 1.73 1.74
0.59 0.32 0.60 0.62 0.42 0.33 0.50 0.20 0.20 Lu 0.18 0.50 0.24 0.27 0.26 0.26 0.27
3.1 1.8 7.7 8.5 4.0 6.3 3.1 0.9 0.4 Hf 0.5 2.5 0.9 8.8 9 9.2 9.6
0.1 0.3 1.5 1.3 0.8 2.1 0.2 <0.1 <0.1 Ta <0.1 0.1 <0.1 6.8 6.9 6.9 7.2
0.8 0.3 1.1 1.1 0.6 2.1 1.8 1.1 1.3 Pb 2.3 0.4 1.3 NA NA NA NA
0.3 0.2 1.5 1.3 0.8 1.8 1.9 0.5 <0.2 Th 0.3 0.5 0.4 9.9 9.9 10.4 10.1
<0.1 <0.1 0.5 0.6 0.4 0.5 0.6 0.2 <0.1 U 0.2 0.1 0.2 15.3 15.4 15.8 16.1

initiation scenario was unlikely (see Section 5.4; see also Section
5.5 for a discussion of different subduction initiation mecha-
nisms). We note that the timeframe of the Whattam and Stern
(2015a) model, which posits subduction initiation as the result
of plume emplacement at 90 Ma, is consistent with the conclu-
sions of Lissinna et al. (2006) who suggest subduction initiation
by 89 Ma.
The active Central American volcanic front stretches ~1100 km
along the western margin of the Caribbean plate from Costa Rica
through Nicaragua, El Salvador and Guatemala to the Guatemala-
Mexico border at the southern margin of the North American
plate (Fig. 1a). The locus of magmatic activity has shifted from east
to west after the Eocene. Early plume- and subduction related se-
quences in the east comprise the ~89-40 Ma Sona-Azuero (Azuero
Marginal Complex of Buchs et al., 2010), 72-69 Ma Golfito, and 70-
39 Ma Chagres-Bayano complexes (Fig. 1b, see also Table 1, Table S1
and the Supplementary Document for description of these com-
plexes). The Sona-Azuero and Chagres-Bayano complexes are
interpreted as being floored by the 89-85 Ma CLIP (Wo €rner et al.,
2009; Buchs et al., 2010; Wegner et al., 2011; Montes et al.,
2012a, 2012b) and the Golfito Complex has been interpreted as
both CLIP oceanic plateau (Hauff et al., 2000) and more recently as
arc (Buchs et al., 2010). Prior to the study of Buchs et al. (2010), it
was generally assumed that most oceanic forearc units along
western Costa Rica and Panama represented uplifted segments of
the CLIP. Similar interpretations have been made for many oceanic
complexes along NW South America and the Leeward Antilles (see
Whattam and Stern, 2015a).
288 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300
3. Methods
3.1. Petrography
Fifty new representative thin sections of the Sona-Azuero
complex lavas were examined under a petrographic microscope.
Sample locations and coordinates are provided in Fig. 2 and
Table S2, respectively, and selected micrographs comprise
Supplementary Fig. S1.
3.2. Elemental analysis
Whole-rock samples (n ¼ 25) of the Azuero complex were
crushed and ground to powders for major, trace and rare earth
element determinations (Table 2). Details of analytical procedures
and elemental analysis are provided in the Supplementary
Document.
3.3. Data compilation and manipulation
3.3.1. Data compilation
For comparison with our 25 new whole-rock geochemical an-
alyses of Azuero lavas (Table 2), we also compile geochemical data
of lavas of the Sona-Azuero (Buchs et al., 2010; Wegner et al., 2011),
Golfito (Hauff et al., 2000; Buchs et al., 2010) and Chagres-Bayano
complexes (Wegner et al., 2011; Montes et al., 2012b). For further
comparison with these Central America forearc complex datasets,
we also compile data of global MORB (Gale et al., 2013) and volcanic
arc basalts (VAB) of depleted arc systems (PetDB, https://www.
earhchem.org/petdb and GEOROC, (http://georoc.mpch-mainz.
gwdg.de/georoc/; see Hawkesworth et al., 1997, for definition of
‘depleted arc systems’); Late Cretaceous to Eocene ‘plume and arc
related’ units (Whattam and Stern, 2015a) in Panama and Costa
Rica; lavas from complexes around the circum-Caribbean region
interpreted as CLIP; SIR ophiolite lavas (Whattam and Stern, 2011);
Izu-Bonin forearc basalts (Reagan et al., 2010; Shervais et al., 2019;
and Izu-Bonin ‘forearc-like’ basalts (Arculus, 2015)). The MORB
dataset (Gale et al., 2013) is mostly basalt (mean of 50.4 wt % SiO2)
but ranges from 44.9 to 70.4 wt % SiO2. Volcanic and plutonic
samples comprising Late Cretaceous to Eocene (~85-39 Ma) ‘plume
and arc-related’ units (Whattam and Stern, 2015a) in Panama and
southernmost Costa Rica are considered as either oceanic plateau
or proto-arc and arc constructed upon the CLIP (e.g., Buchs et al.,
2010) or hybrid plume- and arc-related (Whattam and Stern,
2015a). Basalts interpreted as CLIP (n ¼ 192 tholeiitic and 7
enriched oceanic-island-like basalts) are compiled from GEOROC
and are from the Caribbean Sea; the western margin of the Carib-
bean Plate in Costa Rica (Nicoya, Herradura, Tortugal); the western
margin of NW South America in western Ecuador (Pin ~o

n Formation
and the Pallatanga and Pedernales-Esmeraldas unit) and western
Colombia (Gorgona Island, Serrania de Baudo Formation and the
Western Cordillera); the Lesser Antilles in Aruba (Aruba Lava For-
mation); and the Caribbean Sea (Colombian and Venezuelan
Basins).
3.3.2. Data manipulation
In order to assure that suitable quality analyses of least altered
samples of Sona-Azuero, Golfito and Chagres-Bayano were used in
our study, only samples which yield oxide totals of 96e102 wt %
(excluding loss on ignition, LOI) are plotted. In our geochemical
plots using major elements, oxides are recalculated on an anhy-
drous (volatile-free) basis and normalized to 100%. However, our
geochemical plots and arguments rely primarily on trace elements
known to be non-mobile up to greenschist-facies metamorphic
conditions.
Fig. 3. MgO vs. SiO2 for all lavas of this study superimposed with the fields of MORB
and lavas of depleted arc systems (see text). Abbreviation: enr, enriched.
4. Results
Our lavas are subdivided into three chemical groups on the basis
of chondrite-normalized La/Yb concentrations (see section 4.2.1
and Fig. 5). Groups I, II and III are analogous to those interpreted
by others as plateau basalts, (volumetrically subordinate) enriched
basalts and proto-arc or arc basalts (Buchs et al., 2010; Wegner
et al., 2011), respectively. As shown by Wegner et al. (2011) and
Whattam and Stern (2015a,b), Group I are tholeiitic, Group III range
from tholeiitic to calc-alkaline and Group II are analogous to
oceanic island tholeiites (Supplementary Fig. 2).
4.1. Petrography
Most Group I MORB-like (see Section 5) Azuero lavas are fine-
grained tholeiitic basalts. Textures are typically intergranular to
intersertal or subophitic and mineralogy is primarily plagioclase,
clinopyroxene and alteration products (Supplementary Fig. S1a, b).
In other very fine-grained varieties, clinopyroxene is not obvious
and the mineralogy is dominated by thin, wispy, highly altered
plagioclase laths. Other samples are coarser-grained and some of
these, with large, sericitized plagioclase laths and fresh, subhedral,
interstitial clinopyroxene, are texturally similar to Group III arc-like
(see Section 5) basalts. The crystallization sequence of plagioclase
followed by clinopyroxene is typical of MORB (Bryan, 1983; Pearce
et al., 1984).
Group II oceanic island basalt-like (see Section 5,
Supplementary Fig. S1c, d) lavas are subordinate, making up
~10e20% of the total Azuero samples in our dataset and the com-
bined datasets of Buchs et al. (2010) and Wegner et al. (2011).
Overall, these basalts are moderately to heavily altered and
distinguished by their chocolate brown coloration and ubiquitous
calcite- and chlorite-filled amygdales in the coarser-grained,
weakly porphyritic and medium-grained plagioclase phyric sam-
ples. Fine-grained varieties of these basalts exhibit micro-
porphyritic texture defined by ~1 mm clinopyroxene and
2e5 mm plagioclase phenocrysts. Apatite is an accessory phase
(typically <5%).
Our Group III arc-like (see Section 5) Azuero lavas are rare
(n ¼ 3) and in contrast to lavas of Groups I and II, usually display
micro-porphyritic or unique to this group, robustly porphyritic
textures defined by large (1.5e2.0 mm), commonly euhedral, fresh
 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300 289

Fig. 4. MgO vs. (a) TiO2, (b) Al2O3, (c) FeOt, (d) K2O, (e) Ce, (f) Yb, (g) Y and (h) Zr for all lavas of this study superimposed with the fields of lavas of MORB and depleted arc systems
(basalts only)Abbreviation: enr, enriched.

clinopyroxene phenocrysts (Supplementary Fig. S1e, f). The crys-
tallization sequence, in contrast to the Group I MORB-like lavas, is
clinopyroxene followed by plagioclase which is typical of hydrous
(i.e., arc) magmas in which early plagioclase crystallization is sup-
pressed (Pearce et al., 1984; Cameron 1985; Sisson and Grove 1993).
4.2. Major element chemistry
4.2.1. MgO vs. TiO2, Al2O3, FeOt and K2O
As developed below in Section 4.3, the Sona-Azuero lavas of
Buchs et al. (2010), Wegner et al. (2011) and our dataset are sub-
divided into three groups on the basis of chondrite-normalized La/
Yb concentrations. Group I are interpreted by Buchs et al. (2010)
and Wegner et al. (2011) as plateau basalts; Group II as enriched
basalts (Buchs et al., 2010); and Group III as proto-arc or arc vol-
canics (Buchs et al., 2010; Wegner et al., 2011). These groupings are
referred to below.
A plot of SiO2 vs. MgO (Fig. 3) shows that lavas considered in this
study range from basaltic to rhyolitic with MgO concentrations of
1.06e11.68 wt%. Fractionation patterns for MORB (n ¼ 14, 782) and
290 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300
Fig. 5. Chondrite-normalized REE-normalized plots of Sona-Azuero lavas with 52 or
less wt. SiO2 from this study vs. Sona-Azuero, Golfito and Chagres-Bayano lavas with
less than 52 wt % SiO2 from other studies. The similarity in chondrite-normalized LREE
fractionations, i.e., [La/Yb]CN of our samples with those from other studies (Lissinna,
2005; Buchs et al., 2010; Wegner et al., 2011) in (aec) form the basis for classifica-
tion of our samples into three groups (see Table 3 for further details): (i) Group I, LREE-
depleted, (ii) Group II, LREE-enriched (transitional to alkalic) and (iii) Group III, flat to
slightly LREE-enriched. Chondrite concentrations are from Nakamura (1974).
depleted (see Hawkesworth et al., 1997) arc systems (n ¼ 4, 379) are
broadly similar but with the latter ranging to lower MgO at low SiO2
contents (˂55 wt%) and higher MgO at higher SiO2 (55e72 wt%).
Despite the considerable overlap of MORB and VAB, a large number
of Sona-Azuero Group III arc (n ¼ 8) and Chagre-Bayano samples
(n ¼ 7) exhibit higher arc-like MgO than MORB at silica concen-
trations ˃ 55 wt% SiO2. As well, two Sona-Azuero proto-arc, three
Sona-Azuero Group II samples (including two from this study) and
one Chagres-Bayano sample display lower SiO2 than MORB for a
given MgO concentration at low SiO2 (45e48 wt %) and fall within
the arc domain.
In Fig. 4aed, we plot MgO vs. TiO2, Al2O3, FeOt and K2O for ba-
salts of our Azuero dataset alongside basalts of the Sona-Azuero,
Golfito and Chagres-Bayano suites. For comparison, also plotted
on Fig. 4 are lavas of MOR and depleted arc systems (basalts only).
In all plots there exists considerable overlap between MORB and
VAB but some samples plot uniquely within either field. In
consideration of MgO vs. TiO2, all six Sona-Azuero Type II basalts
plot uniquely within or just above the MORB field and outside of
the arc field. Two of our three Type II Sona-Azuero basalts also have
Al2O3 and FeOt concentrations which plot within the MORB field. In
MgO vs. TiO2 and the remainder of the MgO vs. major oxides plots
(Al2O3, FeOt and K2O), between two to four low-magnesium
(~3e6 wt% MgO) Chagre-Bayano lavas plot uniquely within the
arc field. Similarly, apart from K2O, one low MgO (~4.2 wt% Mg)
Sona-Azuero proto-arc sample plots distinctively within the arc
field; two of our Sona-Azuero Type III arc lavas have TiO2 and FeOt
akin to arc basalts. Additionally, two high MgO (~10e12 wt %)
Chagre-Bayano lavas exhibit MgO vs. TiO2, Al2O3, FeOt and K2O
compositions that plot distinctly within the arc field or on the cusp
between MORB and arc. The remainder of samples plot within the
overlapping area between MORB and VAB.
4.3. Trace element chemistry
4.3.1. MgO vs. Ce, Yb, Y and Zr
We also plot MgO vs. a light and heavy rare earth element (LREE,
Ce and HREE, Yb) and two high field strength elements (Y, Zr) for
basalts of our Azuero dataset alongside basalts of the Sona-Azuero,
Golfito and Chagres-Bayano suites for comparison with MORB and
depleted arc system basalts (Fig. 4eeh). In terms of MgO vs. Ce, two
of our Sona-Azuero high-TiO2 Group II samples plot above both the
MORB and arc fields with the highest Ce of all samples
(46e47 ppm) whereas two Sona-Azuero Group II samples of Buchs
et al. (2009) plot within the MORB field along with one Sona-
Azuero proto-arc and arc sample. Many low-magnesium (˂6 wt%
MgO) basalts plot within the arc field including two of our Sona-
Azuero Group III arc lavas, three Sona-Azuero Group I lavas inter-
preted as plateau by Buchs et al. (2009) and Wegner et al. (2011),
three Sona-Azuero Group III arc basalts and five Chagres-Bayano
basalts. On the other end of the MgO spectrum, one high-MgO
(~11 wt % MgO) Group I Sona-Azuero and one Chagres-Bayano
basalt (~12 wt % MgO) also plot within the arc field. The remain-
ing basalts plot where the MORB and arc fields overlap in MgO vs.
Ce space.
On a plot of MgO vs. Yb (Fig. 4f), four Sona-Azuero Group III arc
basalts including all three of our Sona-Azuero Group III basalts plot
alongside six Chagres-Bayano basalts uniquely within the low-
HREE arc field. Also, three Sona-Azuero Group I, six Sona-Azuero
Group III proto-arc and arc and six Chagre-Bayano basalts plot
outside of both the MORB and arc field; the remaining majority of
Abbreviations: OP, oceanic plateau; PAR, plume- and arc-related. The grey region in (b)
to (f) represent the data presented in (a) and the dashed regions in (d) to (f) represent
the data presented in (c).
 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300 291

samples plot within the MORB realm with Yb ˃ ~1.4.

Many samples exhibit arc-like HFSE concentrations. On a MgO
vs. Y plot (Fig. 4g), our three Sona-Azuero Group III basalts, four of
six Sona-Azuero Group II basalts and eight Sona-Azuero Group III
proto-arc and arc basalts fall alongside almost half of the Chagre-
Bayano basalts (12 of 26 samples) within the arc field. Our three
Sona-Azuero Group III basalts and Chagres-Bayano basalts exhibit
the lowest Y concentrations (˂14 ppm). The remaining basalts fall
within an area of MORB and arc overlap. A similar trend is seen on
MgO vs. Zr (Fig. 4h) with our Sona-Azuero Group III and Chagres-
Bayano basalts exhibiting the lowest Zr arc-like contents of all
samples.

4.3.2. Subdivision of Sona-Azuero lavas into three groups on the

basis of [La/Yb]CN
Table 3 provides the foundation for subdivision of our Azuero
lavas into three groups on the basis of chondrite-normalized La/Yb
concentrations ([La/Yb]CN).
Compositionally, our Azuero Group I basalts are slightly to
moderately light rare earth element depleted ([La/
Yb]CN ¼ 0.75e1.09, mean ¼ 0.96) and compositionally analogous to
the Group I (non-enriched) Azuero ‘plateau’ samples of Buchs et al.
(2010) and the ‘CLIP’ basalts of Wegner et al. (2011) ([La/
Yb]CN ¼ 0.63e1.15, mean ¼ 0.92, Table 3).
Our Group II Azuero basalts are volumetrically subordinate, light
rare earth element-enriched ([La/Yb]CN ¼ 2.29e2.86) and compo-
sitionally similar to the Group II ‘enriched plateau’ samples of Buchs
et al. (2010) ([La/Yb]CN ¼ 1.92e2.81) (Table 3). Taking all enriched
samples into consideration (i.e., of Buchs et al., 2010 and ours),
these basalts are perhaps best described as oceanic island tholeiites
(OIT on Supplementary Figure 2) on the basis of TiO2 vs. Y/Nb re-
lations (Floyd and Winchester, 1975; Winchester and Floyd, 1977).
On the basis of chondrite-normalized rare earth element and N-
MORB-normalized plots (see below), our Group III lavas are VAB-
like and analogous to the Azuero Proto-Arc and Arc samples of
Buchs et al. (2010) and the ‘CLIP’ Arc samples of Wegner et al.
(2011), although our samples range to lower absolute rare earth
element concentrations. Rare earth element patterns of Group III
lavas range from slightly light rare earth element-enriched ([La/
Yb]CN ¼ 1.28e1.58], Table 3) to flat to slightly ‘U-shaped’ (see next
section), i.e., middle rare element-depleted, e.g., two samples have
[La/Gd]CN > 1 and [Yb/Gd]CN > 1).

4.3.3. Chondrite-normalized rare earth element and N-MORB-

normalized incompatible element plots
Chondrite-normalized rare earth element and N-MORB-
normalized plots of Sona-Azuero, Golfito and Chagres-Bayano ba-
salts are provided in Figs. 5 and 6, respectively.

Table 3
Mean and range of La/Yb normalized to chondrite ([La/Yb]CN) of our Group I, II and III
Sona-Azuero lavas vs. Sona-Azuero lavas of other studies (Buchs et al., 2010; Wegner
et al., 2011). Chondrite abundances normalized to are from Sun and McDonough
(1989).

Group mean Our data n mean Others n

[La/Yb]CN [La/Yb]CN
range range

I 0.96 0.75e1.09 12 0.92 0.63e1.15 36

II 2.63 2.29e2.86 3 2.39 1.96e2.81 3
III 1.43 1.28e1.58 3 1.43 0.76e3.21 17

Notes: The Group I of others comprise the Azuero and Sona ‘CLIP’ basalts of Wegner
Fig. 6. N-MORB-normalized plots of Sona-Azuero lavas with 52 or less wt. SiO2 from
et al., (2011) and the Group I Azuero Plateau basalts of Buchs et al., [2010]; the Group
this study vs. Sona-Azuero, Golfito and Chagres-Bayano lavas with less than 52 wt %
II of others comprise the Group II basalts of Buchs et al. (2010); the Group III of
SiO2 from other studies. N-MORB abundances are from Sun and McDonough (1989).
others comprise arc basalts of Azuero, Sona and Coiba Island of Lissinna (2005) and
Abbreviations: (a, e) PAR, plume- and arc-related; (g, i) OP, oceanic plateau. The grey
the basalts of the proto-arc and arc of Buchs et al., (2010). Chondrite abundances are
region in (b) to (f) represent the data presented in (a) and the dashed regions in (d) to
from Sun and McDonough (1989).
(f) represent the data presented in (c).
Table 3 Whattam 2019.
292 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300
Although the rare earth element patterns of Golfito and
Chagres-Bayano basalts are similar to those of Sona-Azuero there
are some significant differences. Rare earth element patterns of
proto-arc lavas of the Golfito Complex (Fig. 5d) generally fall within
the range of the Group I and Group III Sona-Azuero lavas inter-
preted as oceanic plateau and proto-arc/arc respectively, but range
P
to very low absolute rare earth element abundances (mean rare
earth element of 32) and light rare earth element fractionations
(e.g., [La/Yb]CN of 0.81). These values are higher only than that of
P
the Chagres-Bayano Arc (mean rare earth element ~30) and the
lone Chagres-Bayano basalt interpreted as oceanic plateau ([La/
Yb]CN of 0.47) (Fig. 5e and f). Relatedly, this [La/Yb]CN ratio of the
lone Chagres-Bayano basalt interpreted as oceanic plateau (0.47) is
much lower than that of the mean of Sona-Azuero oceanic plateau
and arc basalts (with mean [La/Yb]CN of 0.93 and 1.54, respectively).
Although the Chagres-Bayano Arc exhibits mean [La/Yb]CN inter-
mediate to that of the Sona-Azuero oceanic plateau and proto-arc/
arc (mean [La/Yb]CN of 1.18 vs. 0.93 and 1.54, respectively), the
absolute rare earth element abundances of Chagres-Bayano Arc
P
lavas range to much lower concentrations (mean rare earth el-
ements of 30) of ~ 3 chondrite (Fig. 5f).
In terms of N-MORB-normalized incompatible element pat-
terns, the most conspicuous differences between our Azuero
Groups are: the lack of large ion lithophile element enrichment in
fluid-mobile elements (e.g, Rb, Ba, K, Pb) in Group II, which is well
developed in both Groups I and III; the lack of a negative
Nb-anomaly (with respect to Th and Ce) in Groups I and II, but
which is obvious in Group III; and the pattern of Y and the heavy
rare earth elements (Yb, Lu), which are flat in Groups I and III but
which decrease with decreasing incompatibility in Group II
(Fig. 6aec). The differences between the N-MORB-normalized
patterns of the Sona-Azuero basalts interpreted as CLIP plus our
Group I basalts interpreted as segments of the earliest hybrid,
plume-dominated unit and basalts interpreted as arc are modest,
apart from the prominent Nb depletion and the slightly more, fluid-
modified compositions of the latter. For example, although the CLIP
basalts exhibit no obvious negative Nb-depletion, they do exhibit
modest negative Zr and Ti anomalies and contain only slightly less
Ba and Pb (Fig. 6a).
N-MORB normalized plots of Golfito and Chagres-Bayano basalts
are provided in Fig. 6def. Golfito proto-arc basalts fall completely
within the range of Sona-Azuero oceanic plateau and proto-arc/arc
basalts but range to lower concentrations of large ion lithophile el-
ements (Ba, K, Pb and Sr) (Fig. 6d). Although a negative Nb anomaly is
not well developed in the lone Golfito sample interpreted as oceanic
plateau (Hauff et al., 2000) it does exhibit a subtle negative Zr
anomaly and concentrations of U, Pb, K and Sr higher than or similar
to Golfito basalts interpreted as arc (Buchs et al., 2010); the Golfito
proto-arc samples of Buchs et al. (2010) exhibit subtle but clear
negative Nb anomalies (Fig. 6d)). The N-MORB-normalized incom-
patible element plot of the lone Chagres-Bayano sample interpreted
as oceanic plateau (Fig. 6e) is somewhat anomalous as it lacks the
high LILE enrichments exhibited by most other arc (e.g., Chagres-
Bayano Arc, Fig. 6f) and oceanic plateau samples.
5. Discussion
In the following sections, we explore the significance of our new
geochemical data and that of other workers to address the
following questions: (1) Do Late Cretaceous e Paleogene igneous
rocks of Panama and Costa Rica follow the “subduction initiation
rule” promulgated for the Izu-Bonin-Mariana arc and many
ophiolites?; (2) What is the nature of the contact between “oceanic
plateau” and arc and chemotemporal evolution of magmatism?;
and (3) What was the nature of the early arc system?
5.1. Evidence of adherence of the early central American volcanic
arc system to the subduction initiation rule
We focus here on elements that are not mobilized by hydro-
thermal alteration and metamorphism for the purpose of con-
straining mantle source fertility, degrees of partial melting and
mantle source fugacity. Ratios of Zr/Y, Nb/Y (and Nb/Yb) are useful
as these decrease with mantle depletion and as partial melting
increases. Because of the compatibility of Y and Yb in garnet, these
ratios can also reflect if garnet was present during melt generation
but this may be the case only in Group II enriched oceanic island
tholeiite-like basalts which exhibit low Y and heavy rare earth
element relative to middle rare earth element (Fig. 5b). In Figs. 7e9,
in addition to plotting Panamanian and Costa Rican samples from
this study, we also plot SIR ophiolite basalts and basaltic andesites
for comparison. We do this because SIR ophiolite lavas are (i)
thought to have formed during SI in a proto-forearc environment
and show a distinctive chemotemporal (change in chemistry with
time) trend from less-to more-depleted and subductionemodified
compositions with time (i.e., MORB- to VAB-like) (Whattam and
Stern, 2011) which (ii) along with other chemical criteria (see
below), are critical for understanding the evolution of the mantle
sources for Sona-Azuero and Chagres-Bayano.
5.1.1. Source fertility
In Fig. 7, we plot Zr vs Zr/Y. As demonstrated by Whattam and
Stern (2011) and shown in Fig. 7a, whereas the older lavas of the
earliest-formed lower subduction initiation rule ophiolite units are
mostly MORB-like (84% of the lower unit basalts plot as MORB),
those of the younger, uppermost unit in contrast plot completely
within the VAB field.
These relations show that forearc units progress from tapping/
melting of a less to more depleted source or lower to higher degrees
of partial melting over the course of seafloor spreading during
subduction initiation. Notably, all tholeiiitic Group I Sona-Azuero
basalts interpreted as CLIP (Fig. 7b) plot within the compositional
field defined by all (Group III) Sona-Azuero and most Chagres-
Bayano and Golfito basalts interpreted as proto-arc (Fig. 7c) sug-
gesting a similar source or degrees of partial melting for Sona-
Azuero oceanic plateau and arc. As well, Fig. 7c confirms what
was shown by the N-MORB and rare earth element plots, i.e., that
Chagres-Bayano arc basalts were derived from a more depleted
source, or one that has undergone higher degrees of partial melting
than Sona-Azuero. Similarly, the Golfito basalts exhibit evidence of
derivation from a source intermediate in composition to that of
Sona-Azuero and Chagres-Bayano and Sona-Azuero Group III ba-
salts are more similar in composition to Sona-Azuero Group I than
to Chagres-Bayano. Geochemical relations demonstrate a MORB to
VAB transition from the Sona-Azuero to the Chagres-Bayano unit
and demonstrate adherence to the SIR for the early Central Amer-
ican Volcanic forearc.
5.1.2. Degrees of partial melting
Fig. 7 shows that the source(s) of the Chagres-Bayano basalts
and our Group III Azuero basalts were either more depleted or
underwent higher degrees of partial melting than the Sona-Azuero
source. These plots however, don’t distinguish between the two
possibilities, and if the latter is accurate, provide estimates of the
relative degrees of partial melting.
Cr vs. Y (Fig. 8) can be used to determine relative degrees of
partial melting of basalts (Murton 1989). Superimposed on Fig. 8
are sources 1, 2 and 3 (S1, S2, S3) which represent a MORB
source, a more depleted source after 20% MORB extraction, and a
most depleted source after ~12% extraction of S2, respectively. We
caution that the F values discussed below are model dependent and
we are interested not in absolute degrees of partial melting but
 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300 293

Fig. 7. Zr vs. Zr/Y (Pearce and Norry, 1979) relations of (a) subduction initiation rule
ophiolite (SIRO) basalts, (b) Sona-Azuero, Golfito and Chagres-Bayano basalts inter-
preted as oceanic plateau and our Group I and II Azuero basalts and (c) Sona-Azuero,
Golfito and Chagres-Bayano basalts interpreted as arc and our Group III Azuero basalts.
All Central American samples are basalts only (i.e., no intrusives). Abbreviations in: (a)
bas-and, basaltic andesite; DM, depleted mantle; EM, enriched mantle; WPB, within-
plate basalts. (b) Ha, Hauff et al. (2000). (c) B, Buchs et al. (2010).

differences in partial melting between lavas of Sona-Azuero and

Chagres-Bayano. Also shown in Fig. 8a are fields defined by MORB,
VAB and boninite (Pearce, 1983). Inspection of Fig. 4b reveals that
whereas the lower basalts of subduction initation rule ophiolites
are consistent with ~20e40% partial melting of S1 or 20% partial
melting of S1 and ~10% partial melting of S2, the upper unit basaltic
andesites of subduction initation rule ophiolites indicate higher
degrees of partial melting (~10e40% partial melting of S2). A similar
relation is seen for Sona-Azuero vs. Chagres-Bayano lavas (Fig. 8c
and d). Most Sona-Azuero basalts interpreted as oceanic plateau
plot within the MORB field and are consistent with ~20e40% partial
melting of S1 (Fig. 8c). Approximately half of the Sona-Azuero ba-
salts interpreted as arc and some Chagres-Bayano arc basalts also
plot within the MORB field consistent with ~ up to 40% partial
melting of S1, or perhaps more realistically~ 12% melting of S2
(Fig. 8d). The other half of the Sona-Azuero basalts interpreted as
arc along with the Golfito basalts record partial melting estimates
of 10e20% of S2. Compared to SI ophiolites, the Chagres-Bayano Arc
basalts clearly exhibit evidence of derivation from the highest de-
grees of partial melting. About 35% of the Chagres-Bayano arc ba-
salts plot within the boninite field and are consistent with
derivation via up to about 16e17% partial melting of S3. Although
boninites require >52 wt% SiO2, >8 wt% MgO and <0.5 wt% TiO2
according to the IUGS (Le Bas, 2000), some of these Chagres-Bayano
basalts which plot as boninite contain >8 wt% MgO and <0.5 wt%
TiO2 (e.g., samples P-010223-6, PAN-03-016, Wegner et al., 2011).
Fig. 8. (a) General petrogenetic paths followed by MORB and VAB and fields of MORB,
VAB and boninite (BON) in Y vs. Cr space and Y vs. Cr of (b) subduction initation rule
ophiolite basalts, (c) basalts of the Sona-Azuero, Golfito and Chagres-Bayano com-
plexes interpreted as oceanic plateau and our Groups I and II Azuero basalts and (d) arc
basalts of the Sona-Azuero, Chagres-Bayano and Golfito complexes and our Group III
Azuero basalts (only, no intrusives). All Central American samples are basalts only. The
solid and dashed curves represent incremental batch melting trends and source 1 (S1)
and source 2 (S2) represent source compositions from Murton (1989). S1 represents a
calculated plagioclase lherzolite MORB source comprised of 0.60 oli þ0.20 opx þ0.10
cpx and S2 represents the residue (residual source) subsequent to 20% melt extraction
from S1. The fields of MORB, VAB and within-plate basalt (WPB) are from Pearce
(2008) and the boninite (BON) field is from Dilek et al. (2007). Abbreviations in: (b)
bas-and, basaltic andesite; LU, lower unit; SIRO, subduction initiation rule ophiolite;
UU, upper unit. (c) enr, enriched; Ha, Hauff et al. (2000).
Such high degrees of partial melting reflect exceptionally hot and
shallow forearc mantle beneath young intra-oceanic subduction-
related systems (e.g., Tatsumi and Eggins, 1995). Like the Zr vs Zr/Y
plot (Fig. 7) the Cr/Y plot (Fig. 8) demonstrates an unequivocal
chemotemporal evolution that follows the SIR.
5.1.3. Source oxygen fugacity
It has long been recognized that MORB are distinguishable from
VAB on the basis of Ti/V due to the more oxidized nature (lower Ti/V
294 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300

Fig. 9. Ti/V (Shervais, 1982) of (a) subduction initiation rule ophiolite (SIRO) basalts, (b) Sona-Azuero, Golfito and Chagres-Bayano basalts interpreted as oceanic plateau and our
Groups I and II Azuero basalts and (c) Sona-Azuero, Golfito and Chagres-Bayano basalts interpreted as arc and our Group III basalts. All Central American samples are basalts only
(i.e., no intrusives). Note that two of our enriched Sona-Azuero basalts fall off the plot to the right with 4.7e4.8 wt % TiO2 (~28000e29000 ppm Ti) and ~460 ppm V (Ti/V of ~53) in
Fig. 9b, e within the within-plate basalt field. The Izu Bonin Mariana FAB and Expedition 352 FAB are from Reagan et al. (2010) and Shervais et al. (2019), respectively and the
Expedition 351 FA-like basalts are from Arculus (2015). Abbreviations in: (a) BON, boninite; LU, lower unit, UU, upper unit. (b) alk, alkaline; Ha, Hauff et al. (2000); LU, UU, as in (a).
(c) B, Buchs et al. (2010).

ratio) of arc environments vs. the more reduced nature of MORB
sources (Shervais, 1982). We plot SIR ophiolite lavas vs. the older and
younger segments of the Sona-Azuero, Chagres-Bayano and Golfito
complexes in Ti/V space in Fig. 9aec to demonstrate that the features
developed above in various incompatible and immobile elements
hold true when considering source fugacity; i.e., whereas the older
Sona-Azuero suite (basalts interpreted as oceanic plateau and proto-
arc/arc) is consistent with derivation from a MORB-like source, ba-
salts from Chagres-Bayano on the other hand are clearly more
consistent with derivation from a more arc-like source.
In Fig. 9def, we also plot FAB from the Izu-Bonin Mariana Arc
system (Reagan et al., 2010), Exp. 352 forearc basalts from the outer
Izu-Bonin Mariana Arc system (Shervais et al., 2019), and Exp. 351
Site U1438 (Arculus, 2015) forearc basalt-like lavas (which pres-
ently occupy the rear arc of the Izu-Bonin Mariana Arc system).
Lowermost basalts of the Izu-Bonin Mariana forearc were first
interpreted as trapped Phillipine Sea Plate (DeBari et al., 1999), a
similar interpretation of which was applied to the Tonga forearc
(Meffre et al., 2012). However, these first-formed lavas (forearc
basalt) are MORB-like and instead mark the first magmatic
expression of subduction initiation (Reagan et al., 2010). As shown
in Fig. 9e, the vast majority of the older MORB-like Sona-Azuero
and the lone Chagres-Bayano basalt (all considered as oceanic
plateau) and Golfito basalts plot within the fields defined by the
forearc basalts. Similarly, the vast majority of the younger Chagres-
Bayano Arc basalts also plot within the field defined by forearc
basalts (Fig. 9f), whereby most (77%) of the older Sona-Azuero
basalts interpreted as arc plot outside the fields defined by Izu-
Bonin Mariana (Ishizuka et al., 2006; Reagan et al., 2010) and Exp.
352 forearc basalts (Expedition 352 Scientists, Preliminary Report,
2015) within the MORB field. This again demonstrates a SIR che-
motemporal evolution from early MORB-like lavas to later arc-like
lavas, this time on the basis of oxygen fugacity, in the early Central
American forearc.
5.2. Isotopic constraints on the plume-contaminated nature of the
source
A myriad of studies show the plume-contaminated nature of
Central American forearc lavas as the result of emplacement of the
CLIP which immediately preceded and instigated subduction
initiation (plume-induced subduction initiation, Whattam and
Stern (2015a) and references therein). In terms of source compo-
nents and source mixing, it was shown by Allibon et al. (2008) that
the isotopic compositions of Ecuadorian arc or arc-like lava se-
quences (which according to Whattam and Stern, 2015a are
essentially analogous to arc lavas in Sona-Azuero, Golfito and
Chagres-Bayano) built upon CLIP oceanic plateau fragments could
 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300
be explained in terms of tripolar mixing of three components: a
dominant depleted MORB mantle component; a high m (where m is
a238U/204Pb ratio of an Earth reservoir) component carried by the
CLIP; and a subducted pelagic sediment component. It was further
shown by Allibon et al. (2008) that an average contribution of
70e75% depleted MORB mantle, 12e15% high m, and 7e15% pelagic
sediments (cherts) is required to explain their composition.
Furthermore, Whattam (2018) showed that the Central American
lavas in this study also plot within the 143Nd/144Nd vs 206Pb/204Pb
field defined by the Galapagos Plume at 90 Ma similarly to the
western Late Cretaceous Ecuadorian lavas in the study of Allibon
et al. (2008). It is for these reasons that we use the term ‘plume
contaminated’ to describe the forearc basalts of Sona-Azuero and
the VAB of Chagres-Bayano.
5.3. Nature of the contact between “oceanic plateau” and arc and
chemotemporal evolution of magmatism
Buchs et al. (2010) present a synthetic ‘tectonostratigraphic’
chart (their Fig. 3, reproduced in part in our Fig. 10a) but there is a
~10e12 M.y. hiatus between inferred termination of oceanic
plateau construction at 85 Ma and incipient arc construction at 75-
73 Ma (timing of SI as interpreted by Buchs et al., 2010) at both the
Azuero marginal and Golfito complexes. The only study we know of
that documents the nature of the contact between the lower
“oceanic plateau unit” and the upper arc unit in the Sona-Azuero
Complex is that of Corral et al. (2011) (their Fig. 2, reproduced in
our Fig. 10b). They (Corral et al., 2011) show a conformable contact
between the oceanic plateau basement and the overlying proto-arc
sequence, which in turn is conformably overlain by the arc
sequence (the Rio Quema Formation, see Fig. 10b). This bolsters the
supposition of Whattam and Stern (2015a) of a continuous, unin-
terrupted magmatic succession between underlying lavas inter-
preted as plateau and overlying ones interpreted as proto-arc/arc.
Lissinna et al. (2006) report that earliest arc activity commenced
at 88.3±5 Ma but unfortunately, no further details are available. If
earliest formation of the Panamanian Arc was indeed ~90-89 Ma,
this supports the model of Whattam and Stern (2015a) that plateau
and arc magmatism overlapped beginning about 90 Ma. We also
note that Buchs et al. (2010) remark that some earliest supra-
subduction zone proto-igneous rocks are indistinguishable from
those of ‘typical’ oceanic plateaus which also hints that plateau and
arc magmatism overlapped (or that earliest SI lavas tapped a
strongly plume-contaminated source). Corral et al. (2011) provide a
stratigraphic section though the Rio Quema Formation (Azuero),
showing that the Azuero proto-arc group simply overlies the
Azuero (oceanic plateau) igneous basement which, if the ages of
Lissinna et al. (2006) are accurate, suggest a continuous, uninter-
rupted interval from oceanic plateau to proto-arc/arc magmatism.
This, coupled with the complete compositional overlap between
Sona-Azuero lavas and intrusives interpreted as oceanic plateau
and those interpreted as proto-arc and arc suggest that both are
related in space and time. Clearly, more stratigraphic and
geochronologic investigations are needed but the data in hand
support a model whereby the evolution from earlier CLIP-like ba-
salts to younger arc-like lavas was continuous and uninterrupted
(Fig. 11a) as opposed to a scenario whereby arc construction
occurred upon a pre-existing plateau (Fig. 11b).
It is increasingly clear that the subdivision between oceanic
plateau and arc magmatic sequences in Central America are grada-
tional. We recognize that the subdivision is based on lavas and in-
trusives with negative primitive-mantle normalized high-field
strength element anomalies (particularly Nb) being assigned as arc
and those without as oceanic plateau. However, it was shown by
Whattam and Stern (2011) that the first magmatic expression of
some subduction initation systems is MORB-like, and some basalts
295
of lower unit subduction initiation ophiolites exhibit weak primitive
mantle normalized negative Nb-anomalies, others do not; a robust
negative Nb-anomaly and associated large ion lithophile enrich-
ment is typically seen only in younger lavas and intrusives. We agree
with the interpretation of Whattam and Stern (2015a) that the
earliest magmatic activity at the Central American Arc system in
Panama and Costa Rica was the result of tapping of a hybrid, plume-
subduction source as the result of plume-induced subduction initi-
ation. As shown in Sections 5.1.1.-5.1.3, whereas lavas of both Sona-
Azuero units - i.e., those interpreted as oceanic plateau and those
interpreted as arc - are ‘MORB-like’, the lavas of the Chagres-Bayano
unit are mostly ‘arc-like’. We conclude on this basis that magmatism
evolved continuously from plume-contaminated MORB- to VAB-like
with time consistent with the SIR.
5.4. Subduction initiation mechanisms
As explained by Stern (2004) and later by Stern and Gerya
(2018), subduction initiation can be either spontaneous or
induced. In the case of the former, there are three possible types:
transform collapse, passive margin collapse, and plume head
margin collapse (i.e., plume-induced subduction initiation). In
contrast, induced subduction initiation is triggered by continued
plate convergence following collision, for example of an oceanic
plateau with a subduction zone. There are two types of induced
subduction initiation: a new subduction zone of similar dip to the
original subduction zone can initiate behind the collision (trans-
ference) or in front of the collision and (polarity reversal).
The earliest workers to propose spontaneous subduction initi-
ation at a lithospheric weakness (transform separating older,
denser lithosphere from younger, warmer lithosphere) was
Matsumoto and Tomoda (1983). This idea of spontaneous subduc-
tion initiation along a transform fault was adopted by Stern and
Bloomer (1992) to explain the early evolution of the Izu Bonin
Marianas arc system and by others for other regions of forearc
formation (e.g., Whattam et al., 2006, 2008; Shafaii Moghadam
et al., 2013; Zhou et al., 2018). Other studies have focused on
forced convergence across fracture zones (e.g., Hall et al., 2003).
Subduction nucleation along a passive margin (Marques et al.,
2014; Ueda et al., 2008; Burov and Cloetingh, 2010; Van der Lee
et al., 2008) is implied in the Wilson (1966) cycle but until very
recently (Pandey et al., 2019), there are no known Cenozoic ex-
amples (Stern, 2004). Pandey et al. (2019) recovered forearc basalts
and overlying boninitic-like forearc basalts in the Laxmi Basin,
western Indian passive margin, which record geochemical and
isotopic attributes identical to first formed lavas of the Izu Bonin
Marianas and supra-subduction zone ophiolites and may provide
the first evidence for subduction initiation along a passive margin.
Perhaps the best example of a induced subduction initiation po-
larity reversal event is that caused by collision of the Ontong Java
Plateau, the world’s largest igneous province, which attempted to
subduct to the SW at the Vitiaz Trench ~10-4 Ma (Phinney et al.,
1999; Cooper and Taylor, 1985). Thick, buoyant lithosphere of the
Plateau jammed the subduction zone and caused a subduction
polarity flip to the NE. As explained in the Introduction, Buchs et al.
(2010) suggested that subduction initiation at the Central American
Arc was tectonically induced and resulted from compression of the
thickened Caribbean Plate during westward migration of the
Americas and collision of the thickened Caribbean Plate with South
America. As explained above in Section 5.3, on the basis of a
comprehensive set of geochemical, isotopic and geochronologic
data and tectonic considerations, we concur with Whattam and
Stern (2015a, 2015b) that arc construction in Central America
began instead about 90 Ma as the result of plume-induced sub-
duction initiation. A second example of plume-induced subduction
initiation has been recently proposed for the Cascadia subduction
296 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300

Fig. 10. (a) Chemotemporal relations between the Golfito, Sona-Azuero and Chagres-Bayano as shown on the ‘synthetic tectonostratigraphic chart’ of Buchs et al. (2010). Unit
acronyms (from left to right): AP, Azuero (oceanic) Plateau (89-85 Ma); GA, Golfito Arc (75-66 Ma); SA, Sona-Azuero Arc (75-39 Ma); CBA, Chagres-Bayano Arc (70-39 Ma). (b)
Stratigraphy of an Azueru section (the Rio Quema Formation) reproduced from Corral et al. (2011) which provides relations between the Azuero oceanic ‘plateau’, proto-arc and arc.
Other acronyms: PISI, plume-induced subduction initiation. References for ages are provided in Table 1.

zone in the Pacific NW, USA (Stern and Dumitru, 2019) and plume-
induced subduction initiation has shown to be viable on the basis of
numerical thermomechanical modelling (Gerya et al., 2015).
5.5. Physical nature of the early arc system and tectonic
considerations
In their study of Panamanian magmatism circa 75 Ma to present,
Wegner et al. (2011) suggest that arc magmatism began in the Sona
and Azuero peninsulas at ~71 Ma (Sona-Azuero Arc), then shifted to
the Chagres-Bayano region at 66 Ma; Wegner et al. (2011) also
apparently consider these as two distinct arcs. We disagree with
these notions for several reasons. Firstly, arc lengths are typically
not at the 100s of km scale but rather 1000s of km scale. For
example, the modern-day Central American Volcanic Arc system is
~1500 km long and the Izu-Bonin Marianas Arc system is ~2800 km
 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300 297

Fig. 11. Two alternative tectonic models to reconcile subduction initiation at the Central American Volcanic Arc system. (a) The earliest stages of subduction initiation were
catalyzed by plume emplacement (modified from Whattam and Stern, 2015a). This event began to first form the Sona-Azuero complex which was then followed by incipient
formation of the Chagres-Bayano complex, subsequent to slab rollback and hinge retreat. (b) Collision-induced subduction initiation (see Buchs et al., 2010 for details) and sub-
sequent arc construction upon a ~10 m.y. oceanic plateau (based on the ideas of Buchs et al., 2010 and Wegner et al., 2011).

long. The magmatic arc ‘belts’ along the Sona-Azuero peninsula and
the Chagres-Bayano region are only ~200e250 km long, but
restoration of paleomagnetically derived vertical-axis rotations
(Montes et al., 2012a) suggests that the combined length of the
Campanian to Eocene Central American arc was at least 1000 km.
Furthermore, combined geochemical and geochronological evi-
dence suggests that the nascent Central American arc stretched
westwards from the Nicoya peninsula in western Costa Rica some
200 km south of Nicaragua, to either present-day western Ecuador
or Colombia some 1400 km to the east (Whattam and Stern, 2015a).
Secondly, the age constraints of Lissinna (2005) indicate that Sona-
Azuero magmatic activity was contemporaneous with Chagres-
Bayano magmatism between 70 and 40 Ma. Thirdly, unequivocal
‘arc’ magmatism at the Sona-Azuero and Golfito complexes began
by at least 75 Ma (Buchs et al., 2010) but the 89-85 Ma Sona-Azuero
CLIP MORB-like oceanic plateau basalts are largely indistinguish-
able from the younger Sona-Azuero arc and proto-arc basalts.
Furthermore, Lissinna et al. (2006) suggest arc magmatism began
by ~90 Ma. Finally, as shown by Whattam and Stern (2015a), most
(~90-85 Ma) oceanic plateau basalts from the western edge of the
CLIP (Costa Rica, e.g., Nicoya, Herradura, Tortugal) including those
of Sona-Azuero and the NW segment of the South American Plate in
Ecuador and Western Colombia, exhibit subduction modified
compositions. Although a lateral west to east transition in chem-
istry cannot be categorically ruled out, the relations in the chem-
istry of Sona-Azuero vs. Chagres-Bayano basalts are more
consistent with a vertical/stratigraphic transition, although erup-
tions may have become more arc-like as magmatism retreated from
the trench. This chemostratigraphic progression can also be
considered as chemotemporal as the change in magmatic affinities
is largely vertical and hence a record also of change in chemistry
with time; this chemotemporality is identical to that recorded by
the slightly older Sona-Azuero and the younger Chagres-Bayano
complexes. Thus, we suggest that the Chagres-Bayano arc
segment represents a vertical as well as lateral migration of arc
magmatism that is also seen for Sona-Azuero circa 71 Ma (Wegner
et al., 2011) (Fig. 7). In other words, the Chagres-Bayano arc
segment may represent the carapace of the early established Cen-
tral American Arc system.
If Chagres-Bayano does indeed represent the latest phase of
early Central American Arc magmatism, the present position of the
Chagres-Bayano arc segment indicates that it is offset to the north.
In the tectonic model of Whattam et al. (2012), the early (pre-
40 Ma) Central American Arc was shut down at about 40 Ma as the
result of collision of an oceanic plateau with the arc system (see also
Kerr and Tarney, 2005). As explained by Montes et al. (2012a,
2012b) and Whattam et al. (2012), this collision caused left-lateral
offset of the originally contiguous early (pre-40 Ma) Central
American Volcanic Arc system which stretched from at least
western Costa Rica in the NW to western Ecuador or Colombia in
the SE (Whattam and Stern; 2015a) and resulted in a >100 km
displacement to the north between 40 and 30 Ma (Lissinna, 2005;
Montes et al., 2012a, 2012b). Alternatively, left-lateral offset may
have been caused by a shift in the subducted plateau from Azuero to
a region to the east (Lissinna et al., 2006) or as the result of collision
of South America with the Panamanian Arc system (Farris et al.,
2011; Barat et al., 2014). Regardless of the mechanism of displace-
ment or the nature of the colliding body, we suggest that the
Chagres-Bayano arc segment was translated to the north after
originally forming the carapace and rear of the early arc system
subsequent to slab rollback but prior to displacement (Fig. 11).
6. Conclusions
An unambiguous chemotemporal transition from early-formed
plume-contaminated MORB-like to VAB-like lavas is seen in Late
Cretaceous-Paleogene magmatic arc sequences in Panama as
revealed by incompatible trace element compositions which illus-
trate fundamental differences between the older Sona-Azuero
complex and the younger Chagres-Bayano complex. While the
298 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300
trace element signatures of basalts from both units are similar, the
absolute REE concentrations of Chagres-Bayano basalts range to
much lower S rare earth element concentrations (12e52 ppm) than
those of Sona-Azuero (S rare earth elements ¼ 33e102 ppm),
exhibit more negative high field strength element-anomalies and
higher fluid mobile element/immobile element ratios. Relative to
Sona-Azuero, the 70-39 Ma Chagres-Bayano basalts reflect higher
degrees of partial melting of a more depleted and oxidized mantle
source that received greater slab-derived shallow and deep sub-
duction contributions.
Moreover, the Sona-Azuero basalts interpreted as proto-arc and
arc are compositionally more similar to the Sona-Azuero basalts
interpreted as oceanic plateau than they are to the Chagres-Bayano
basalts, which if age estimates are accurate suggest that CAVAS
subduction began by ~89 Ma. The overarching chemotemporal
trend recorded by Sona-Azuero and Chagres-Bayano sequences
may reflect the combined effects of a west to east migration of
magmatism at 68 Ma as well as a vertical/stratigraphic transition.
We suggest that the arc segment now preserved at Chagres-Bayano
was originally generated above and to the rear of Sona-Azuero
forearc magmatism and was displaced to the north perhaps due
to collision with an oceanic plateau on the Cocos plate, which shut
down the early arc system ~40 Ma.
The chemotemporal progression from older MORB-like lavas
and intrusives at Sona-Azuero to younger VAB-like lavas at
Chagres-Bayano to the east confirms adherence of the Late Creta-
ceous to Eocene Central American forearc to the SIR, similar to
Tethyan ophiolites and the Izu-Bonin-Mariana forearc. Importantly,
we note that this adherence is independent of the two competing
tectonic models described. In other words, even if arc construction
occurred intrusive into and atop a pre-existing oceanic plateau and
we exclude the MORB-like affinities of the oceanic plateau, the SIR
still applies as the Sona-Azuero Proto-Arc and arc lavas and in-
trusives are also MORB-like in contrast to those of the Chagres-
Bayano Arc which instead exhibit unequivocal volcanic arc
affinities.
Acknowledgements
We are grateful to Mark Tupper, NSF EAR-0824299, National
Geographic, Smithsonian Institution, and Ricardo Perez S.A. which
partially funded this project. S. Zapata and D. Ramirez are
acknowledged for their help in the field and for generating the base
map with the sample localities. Access to field areas and collection
permits were granted by Ministerio de Industria y Comercio of
Panam
€rg Geldmacher, Jeffrey Ryan, and Yamirka
a. We thank Jo
Rojas-Agramonte for their reviews of a previous version of the
manuscript and Fernando Marques, Kosuke Ueda and an anony-
mous reviewer for their reviews of the current manuscript and
positive comments which resulted in a greatly improved paper.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.gr.2019.10.002.
References
Alabaster, T., Pearce, J.A., Malpas, J., 1982. The volcanic stratigraphy and petrogenesis
of the Oman ophiolite complex. Contrib. Mineral. Petrol. 81, 168e183.
Allibon, J., Monjoie, P., Lapierre, H., Jaillard, E., Bussy, F., Bosch, D., Senebier, F., 2008.
The contribution of the young Cretaceous Caribbean Oceanic Plateau to the
genesis of late Cretaceous arc magmatism in the Cordillera Occidental of
Ecuador. J. South Am. Earth Sci. 26, 355e368.
Arculus, R.J., and the 29 others of IODP Expedition 351, 2015. A record of subduction
initiation 800 in the Izu-Bonin-Mariana arc. Nature Geoscience. https://doi.org/
10.1038/ngeo2515.

pinay, B., Sosson, M., Müller, C., Baumgartner, P.,
Barat, F., Mercier de Le
Baumgartner-Mora, C., 2014. Transition from the Farallon Plate subduction to
collision between south and Central America: geological evolution of the
Panama isthmus. Tectonophysics 622, 145e167.
Bryan, W.B., 1983. Systematics of modal phenocryst assemblages in submarine
basalts: Petrologic implications. Contrib. Mineral. Petrol. 83, 62e74.
Buchs, D.M., Arculus, R.J., Baumgartner, P.O., Baumgartner-Mora, C., Ulianov, A.,
2010. Late cretaceous arc development on the SW margin of the Caribbean
plate: insights from the Golfito, Costa Rica, and Azuero, Panama, complex.
Geochem. Geophys. Geosyst. 11 https://doi.org/10.1029/2009GC002901.
Buchs, D.M., Baumgartner, P.O., Baumgartner-Mora, C., Bandini, A.N., Jackett, S.-J.,
Diserens, M.-O., Stucki, J., 2009. Late Cretaceous to Miocene seamount accretion
and me
lange formation in the Osa and Burica peninsulas (southern Costa Rica):
episodic growth of a convergent margin. In: James, K.H., Lorente, M.A.,
Pindell, J.L. (Eds.), The Origin and Evolution of the Caribbean Plate, vol. 328.
Geological Society Special Publication, pp. 411e456.
Burov, E., Cloetingh, S., 2010. Plume-like upper mantle instabilities drive subduction
initiation. Geophys. Res. Lett. 37, L03309. https://doi.org/10.1029/
2009GL041535.
Cameron, W.E., 1985. Petrology and origin of primitive lavas from the Trodoos
ophiolite, Cyprus. Contrib. Mineral. Petrol. 89, 239e255. https://doi.org/
10.1007/BF00379457.
Cooper, P.A., Taylor, B., 1985. Polarity reversal in the Solomon Islands arc. Nature
314, 428e430.

mez-Gras, D., Corbella, M., Canals, A.,  Pineda-Falconett, M.,
Corral, I., Griera, A., Go
Cardellach, E., 2011. Geology of the cerra Quema Au-Cu deposit (Azuero
peninsula, Panama). Geol. Acta 9 (3e4), 481e498, 1 0 . 1 3 4 4/1 0 5 . 0 0 0 0 0 1 7
4 2.
DeBari, S.M., Taylor, B., Spencer, K., Fujioka, K., 1999. A trapped Philippine Sea plate
origin for MORB from the inner slope of the Izu-Bonin trench. Earth Planet. Sci.
Lett. 174, 183e197.
Dengo, G., 1962. Tectonic-igneous sequence in Costa Rica. In: Engel, A.E.J.,
James, H.L., Leonard, B.F. (Eds.), Petrologic Studies: A Volume in Honor of A. F.
Buddington. Geological Society of America, New York, pp. 133e161.
Di Marco, G., 1994. Les terrains accre
te


s du Costa Rica: Evolution teconostrati-
graphique de la marge occidentale de la Plaque Caraïbe. Mem. Geol., vol. 20
Univ. de Lausanne, Lausanne, Switzerland, p. 20.
Di Marco, G., Baumgartner, P.O., Channell, J.E.T., 1995. Late Cretaceous-early Tertiary
paleomagnetic data and a revised tectonostratigraphic subdivision of Costa Rica
and western Panama. In: Mann, P. (Ed.), Geologic and Tectonic Development of
the Caribbean Plate Boundary in Southern Central America, 295, pp. 1e27. Spec.
Pap. Geol. Soc. Am.
Dilek, Y., Furnes, H., Shallo, M., 2007. Suprasubduction zone ophiolite formation
along the periphery of Mesozoic Gondwana. Gondwana Res. 11, 453e475.
Dilek, Y., Furnes, H., Shallo, M., 2008. Geochemistry of the Jurassic Mirdita Ophiolite
(Albania) and the MORB to SSZ evolution of a marginal basin oceanic crust.
Lithos 100, 174e209.
Gale, A., Dalton, C.A., Langmuir, C.H., Su, Y., Schilling, J.-G., 2013. The mean
composition of ocean ridge basalts. Geochem. Geophys. Geosyst. 14, 489e518.
https://doi.org/10.1029/2012GC00433.
Floyd, P.A., Winchester, J.A., 1975. Magma Type and Tectonic Setting Discrimination
Using Immobile Elements. Earth and Planetary Science Letters 27, 211e218.
https://doi.org/10.1016/0012-821X(75)90031-X.
Gazel, E., Carr, M.J., Hoernle, K., Feigenson, M.D., Hauff, F., Szymanski, D., Van Den
Bogaard, P., 2009. The Galapagos-OIB signature in southern Central America:
mantle re-fertilization by arc-hotspot interaction. Geochem. Geophys. Geosyst.
Q02S11. https://doi.org/10.1029/2008GC002246.
Gazel, E., Hayes, J.L., Hoernle, K., Kelemen, P., Everson, E., Holbrook, W.S., Hauff, F.,
Van Den Bogaard, P., Vance, E.A., Chu, S., Calvert, A.J., Carr, M.J.,
Yogodzinski, G.M., 2015. Continental Crust generated in oceanic arcs. Nat.
Geosci. 8, 321e327. https://doi.org/10.1038/ngeo2392.
Geldmacher, J., Hanan, B.B., Blichert-Toft, J., Haarp, K., Hoernle, K., Hauff, F., 2001.
Hafnium isotope variations in volcanic rocks from the Caribbean Large Igneous
Province and Gala
pagos hotspot tracks. Geochem. Geophys. Geosyst. 4 https://
doi.org/10.1029/2002GC000477.
Geldmacher, J., Hoernle, K., Van Den Bogaard, P., Hauff, F., Klügel, A., 2008. Age and
geochemistry of the Central American forearc basement (DSDP Leg 67 and 84):
insights into Mesozoic arc volcanism and seamount accretion on the fringe of
the Caribbean LIP. J. Petrol. 49, 1781e1815. https://doi.org/10.1093/petrology/
egn046.
Gerya, T., Stern, R.J., Baes, M., Sobolev, S.V., Whattam, S.A., 2015. Plate tectonics on
the Earth triggered by plume-induced subduction initiation. Nature 527,
221e225. https://doi.org/10.1038/nature15752.
Giudice, D., Recchi, G., 1969. Geología del area del projecto minero de Azuero.
Ejectora, informe tecnico. Gobierno de la Repub. de Panama
, Panama City, p. 48.
Hall, C.E., Gurnis, Sdrolias, M., Lavier, L.L., Muellar, R.D., 2003. Catastrophic initiation
of subduction following forced convergence across fracture zones. Earth Planet.
Sci. Lett. 212, 15e30.
Hauff, F., Hoernle, K., Schmincke, H.eU., Werner, R., 1997. A Mid-Cretaceous origin
for the Galapagos hotspot: volcanological, petrological and geochemical evi-
dence from Costa Rican oceanic crustal fragments. Geol. Rundsch. 86, 141e155.
Hauff, F., Hoernle, K.A., van den Bogaard, P., Alvarado, G.E., Garbe-Schonberg, D.,
2000. Age and geochemistry of basaltic complexes in western Costa Rica:
contributions to the geotectonic evolution of Central America. Geochem. Geo-
phys. Geosyst. 1. Paper No. 1999GC000020.
Hawkesworth, C.J., Turner, S.P., McDermott, F., Peate, D.W., van Calsteren, P., 1997. U-
Th isotopes in arc magmas: implications for element transfer from the
 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300
subducted crust. Science 276, 551e555.
Hoernle, K., van den Bogaard, P., Werner, R., Lissinna, B., Hauff, F., Alvarado, G.,
Garbe-Schonberg, D., 2002. Missing history (16e71 Ma) of the Galapagos hot-
spot: implications for the tectonic and biological evolution of the Americas.
Geology 30, 795e798. https://doi.org/10.1130/0091-7613(2002).
Ishizuka, O., Kimura, J.-I., Li, Yi B., Stern, R.J., Reagan, M.K., Taylor, R.N., Ohara, Y.,
Bloomer, S.H., Ishii, T., Hargrove III, U.S., Haraguchi, S., 2006. Early stages in the
evolution of the Izu-Bonin arc volcanism: new age, chemical and isotopic
constraints. Earth Planet. Sci. Lett. 250, 385e401.
Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, S., Harigane, Y.,
Sakamoto, I., Miyajima, Y., Yuasa, M., Dunkley, D.J., 2011. The timescales of
subduction initiation and subsequent evolution of an oceanic island arc. Earth
Planet. Sci. Lett. 306, 229e240.
Kerr, A.C., Tarney, J., 2005. Tectonic evolution of the Caribbean and northwestern
South America: The case for accretion of two Late Cretaceous oceanic plateaus.
Geology 33, 269e272. https://doi.org/10.1130/G21109.1.
Kerr, A.C., White, R.V., Saunders, A.D., 2000. LIP reading: recognizing oceanic
plateaux in the geological record. J. Petrol. 41 (7), 1041e1056.
Kerr, A.C., White, R.V., Thompson, P.M., Tamez, J., Saunders, A.O., 2003. In:
Bartolini, C., Bufler, R.T., Blickwede, J. (Eds.), The Circum-Gulf of Mexico and the
CaribbeanL Hydrocarbon, Habitats, Basin Formation and Plate Tectonics:
American Associate of Petroleum Geologists (MPG) Memoir 79, pp. 126e188.
Kolarsky, R.A., Mann, P., Monechi, S., Meyerhoff, H.D., Pessagno Jr., E.A., 1995.
Stratigraphic development of southwestern Panama as determined from inte-
gration of marine seismic data and onshore geology. Special Paper. In: Mann, P.
(Ed.), Geologic and Tectonic Development of the Caribbean Plate Boundary in
Southern Central America, vol. 295. Geological Society of America, pp. 159e200.
Lissinna, B., Hoernle, K., van den Bogaard, P., 2002. Northern migration of arc
volcanism in western Panama: evidence for subduction erosion? Eos Trans-
actions, AGU 83, 1463e1464.
Le Bas, M.J., 2000. IUGS reclassification of the high-Mg and picritic volcanic rocks.
J. Petrol. 41, 1467e1470.
Lissinna, B., 2005. A Profile Though the Central American Landbridge in Western
Panama: 115 Ma Interplay between the Gal
apagos Hotspot and the Central
American Subduction Zone. Ph.D. thesis. Christian-Albrechts University, Kiel,
Germany, p. 102.
Lissinna, B., Hoernle, K., Hauff, F., van den Bogaard, P., Sadofsky, S., 2006. The
Panamanian island arc and Gala
pagos hotspot: a case study for the long-term
evolution of arc/hotspot interaction. Geophys. Res. Abstr. 8, 05106.
Marques, F.O., Cabral, F.R., Gerya, T.V., Zhu, G., May, D.A., 2014. Subduction initiates
at straight passive margins. Geology 42, 331e334. https://doi.org/10.1130/
G35246.1.
Matusomoto, T., Tomoda, Y., 1983. Numerical simulation of the initiation of sub-
duction at the fracture. J. Phys. Earth 31, 183e194.
Maury, R.C., Defant, M.J., Bellon, H.,, de Boer, Z.J., Stewart, R.H., Cotten, J., 1995. Early
Tertiary arc volcanics from eastern Panama. In: Mann, P. (Ed.), Geologic and
Tectonic Development of the Caribbean Plate Boundary in Southern Central
America, vol. 295. Specical Paper of the Geological Society of America,
pp. 29e34.
Meffre, S., Falloon, T.J., Crawford, T.J., Hoernle, K., Hauff, F., Duncan, R.A.,
Bloomer, S.H., Wright, D.J., 2012. Basalts erupted along the Tongan fore arc
during subduction initiation: evidence from geochronology of dredged rocks
from the Tonga fore arc and trench. Geochem. Geophys. Geosyst. 13, Q12003.
https://doi.org/10.1029/2012GC004335.
Metti, A., Recchi, G., 1976. Geología de la Peninsula de Son
a e Isla de Coiba. In:
Memoria Segundo Congreso Latinoamericano de Geología, vol. 2, pp. 541e553
(Sucre, Caracas).
Montes, C., Bayona, G.A., Cardona, A., Buchs, D.M., Silva, C.A., Moro
n, S., Hoyos, N.,
Ramírez, D.A., Jaramillo, C.A., Valencia, V., 2012a. Arc-continent collision and
orocline formation: closing of the central American seaway. J. Geophys. Res. 117,
B04105. https://doi.org/10.1029/2011JB008959.
Montes, C., Cardona, A., McFadden, R., Moro
n, S.E., Silva, C.A., Restrepo-Morono, S.,
Ramírez, D.A., Hoyos, N., Wilson, J., Farris, D., Bayona, G.A., Jaramillo, C.A.,
Valencia, V., Bryan, J., Flores, J.A., 2012b. Evidence for middle Eocene and
younger land emergence in central Panama: implications for isthmus closure.
Geol. Soc. Am. Bull. https://doi.org/10.1130/B30528.1.
Murton, B.J., M.J., 1989. Tectonic controls on boninite genesis. In: Saunders, A.D.
(Ed.), Magmatism in the ocean basins, 43. Geological Society of London Special
Publication, pp. 347e377.
Nakamura, N., 1974. Determination of REE, Ba, Fe, Mg, Na and K in carbonaceous and
ordinary meteorites. Geochem. Cosmochim. Acta 38, 757e775.
Obando, J.A., 1986. Sedimentología y tecto
nica del Creta
cico y Paleo
geno de la re-
gion de Golfito, Península de Burica y Península de Osa, Provincia de Puntar-
enas, Costa Rica. licenciatura thesis. Univ. de Costa Rica, San Jose
, Costa Rica,
p. 211. Escuela Centroam. de Geol.
Pandey, D.K., Pandey, A., Whattam, S.A., 2019. Relict subduction initiation along a
passive margin in the northwest Indian Ocean. Nat. Commun. 10, 2248. https://
doi.org/10.1038/s41467-019-10227-8.
Pearce, J.A., 1983. Role of the subcontinental lithosphere in magma genesis at active
continental margins. In: Hawkesworth, C.J., Norry, M.J. (Eds.), Continental Ba-
salts and Mantle Xenoliths, pp. 230e249 (Nantwich, Shiva).
Pearce, J.A., 2003. Supra-subduction zone ophiolites: the search for modern ana-
logues. In: Dilek, Y., Newcomb, S. (Eds.), Ophiolite Concept and the Evolution of
Geological Thought, vol. 373. Geological Society of America Special Paper,
pp. 269e293.
Pearce, J.A., 2008. Geochemical fingerprinting of oceanic basalts with applications
299
to ophiolite classification and the search for Archean oceanic crust. Lithos 100,
14e48.
Pearce, J.A., Lippard, S.J., Roberts, S., 1984. Characteristics and tectonic significance
of supra-subduction zone ophiolites. In: Kokelaar, P., Howels (Eds.), Geology of
Marginal Basins, vol. 16. Geological Society of London Special Publication,
pp. 77e94.
Pearce, J.A., Norry, N.J., 1979. Petrogenetic implications of Ti, Zr, Y, and Nb variations
in volcanic rocks. Contrib. Mineral. Petrol. 69, 33e49.
Pearce, J.A., Robinson, P.T., 2010. The Troodos ophiolitic complex probably formed in
a subduction initiation, slab edge setting. Gondwana Res. 18, 60e81.
Phinney, E.J., Mann, P., Coffin, M.F., Shipley, T.H., 1999. Sequence stratigraphy,
structure, and tectonic history of the southwestern Ontong Java Plateau adja-
cent to the North Solomon Trench and Solomon Islands arc. J. Geophys. Res. 104
(B9), 20446e20449.
Pindell, J.L., Kennan, L., 2009. Tectonic evolution of the Gulf of Mexico, Caribbean
and northern South America in the mantle reference frame: an update. In:
James, K.H., Lorente, M.A., Pindell, J.L. (Eds.), The Origin and Evolution of the
Caribbean Plate, vol. 328. Geological Society, London, pp. 1e55.
Reagan, M.K., Ishizuka, O., Stern, R.J., Kelley, K.A., Ohara, Y., Blichert-Toft, J.,,
Bloomer, S.eH., Cash, J., Fryer, P., Hanan, B.B., Hickey-Vargas, R., Ishii, T.,
Kimura, J.I., Peate, D.W., Rowe, M.C., Woods, M., 2010. Fore-arc basalts and
subduction initiation in the Izu-Bonin-Mariana system. Geochem. Geophys.
Geosyst. 11, Q03X12. https://doi.org/10.1029/2009GC002871.
Reagan, M.K., Pearce, J.A., Petronotis, K., Almeev, R.R., Avery, A.J., Carvallo, C.,
Chapman, T., Christeson, G.L., Ferre
, E.C., Godard, M., Heaton, D.E.,
Kirchenbaur, M., Kurz, W., Kutterolf, S., Li, H., Li, Y., Mishibayashi, K., Morgan, S.,
Nelson, W.R., Prytulak, J., Python, M., Robertson, A.H.F., Ryan, J.G., Sager, W.W.,
Sakuyama, T., Shervais, J.W., Shimizu, K., Whattam, S.A., 2017. Subduction
initiation and ophiolite crust: new insights from IODP drilling. Int. Geol. Rev.
https://doi.org/10.1080/00206814.2016.1276482.
Robertson, A., 2004. Development of concepts concerning the genesis and
emplacement of Tethyan ophiolites in the Eastern Mediterranean and Oman
regions. Earth Sci. Rev. 66, 331e387.
Saccani, E., Photiades, A., 2004. Mid-ocean ridge and supra-subduction affinities in
the Pindos Massif ophiolites (Greece): implications for magma genesis in a
proto-forearc setting. Lithos 73, 229e253.
Schmidt-Effing, R., 1979. Alter und Genese des Nicoya-Komplexes, einer ozeani-
schen Palaeokruste (Oberjura bis Eozaen) im suedlichen Zentralamerika. Geol.
Rundsch. 68, 457e494. https://doi.org/10.1007/BF01820803.
Shafaii Moghadam, H., Corfu, F., Stern, R.J., 2013. U-Pb zircon ages of Late Cretaceous
Nain-Dehshir ophiolites, central Iran. J. Geol. Soc. 170, 175e184. https://doi.org/
10.1144/jgs2012066.
Shervais, J.W., 1982. Ti-V plots and the petrogenesis of modern and ophiolitic lavas.
Earth Planet. Sci. Lett. 59, 110e118.
Shervais, J.W., Reagan, M., Haugen, E., Almeev, R.R., Pearce, J.A., Prytulak, J.,
Ryan, J.G., Whattam, S.A., Godard, M., Chapman, T., Li, H., Kurz, W., Nelson, W.R.,
Heaton, D., Kirchenbaur, M., Shimizu, K., Sakuyama, T., Li, Y., Vetter, S.K., 2019.
Magmatic response to subduction initiation: Part 1. Fore-Arc basalts of the izu-
bonin arc from IODP expedition 352. Geochem. Geophys. Geosyst. 20 https://
doi.org/10.1029/2018GC007731.
Sinton, C.W., Duncan, R.A., Denyer, P., 1997. Nicoya Peninsula, Costa Rica: a single
suite of Caribbean oceanic plateau magmas. J. Geophys. Res. 102 (15) https://
doi.org/10.1029/97JB00681, 507-515,520.
Stern, R.J., 2004. Subduction initiation: spontaneous and induced. Earth Planet. Sci.
Lett. 226, 275e292.
Stern, R.J., Bloomer, S.H., 1992. Subduction zone infancy: examples from the Eocene
izueboninemariana and jurassic California. Geol. Soc. Am. Bull. 104, 1621e1636.
Stern, R.J., Dumitru, T.A., 2019. Eocene initiation of the Cascadia subduction zone: a
second example of plume-induced subduction initiation? Geosphere. https://
doi.org/10.1130/GES02050.1.
Stern, R.J., Gerya, T., 2018. Subduction initiation in nature and models: a review.
Tectonophysics 746, 173e198.
Sisson, T.W., Grove, T.L., 1993. Experimental investigations of the role of H2O in calc-
alkaline differentiation and subduction zone magmatism. Contributions to
Mineralogy and Petrology 113, 143e166. https://doi.org/10.1007/BF00283225.
Stern, R.J., Reagan, M., Ishizuka, O., Ohara, Y., Whattam, S., 2012. To understand
subduction initiation, study forearc crust: to understand forearc crust, study
pphiolites. Lithosphere. https://doi.org/10.1130/L183.1.
Sun, S.S., McDonough, W.F., 1989. Chemical and isotopic systematic of oceanic ba-
salts: implications for mantle composition and processes. In: Saunders, A.D.,
Norry, M.J. (Eds.), Magmatism in the Ocean Basins, vol. 42. Geological Society of
London Special Publication, pp. 313e345.
Tatsumi, Y., Eggins, S., 1995. Subduction Zone Magmatism. Blackwell, London, p. 211.
Tomascak, P.B., Ryan, J.G., Defant, M.J., 2000. Lithium isotope evidence for light
element decoupling in the Panama subarc mantle. Geology 28, 507e510.

, L., Proenza, J.A., Marchesi, C., Garcia-Casco, A., Lewis, J.F., 2017. Petrogenesis of
Torro
meta-volcanic rocks from the maimo
n formation (Dominican republic):
geochemical record of the nascent greater Antilles paleo-arc. Lithos 278e281,
255e273. https://doi.org/10.1016/j.lithos.2017.01.031.
Ueda, K., Gerya, T., Sobolev, S.V., 2008. Subduction initiation by thermal-chemical
plumes: numerical studies. Phys. Earth Planet. Inter. 171, 296e312. https://
doi.org/10.1016/j.pepi.2008.06.032.
Van der Lee, S., Regenauer-Lieb, K., Yuen, D.A., 2008. The role of water in connecting
past and future episodes of subduction. Earth Planet. Sci. Lett. 273, 15e27.
https://doi.org/10.1016/j.epsl.2008.04.041.
Wegner, W., Wo €rner, G., Harmon, R.S., Jicha, B.R., 2011. Magmatic history and
300 S.A. Whattam et al. / Gondwana Research 79 (2020) 283e300
evolution of the Central American land bridge in Panama since Cretaceous
times. Geol. Soc. Am. Bull. 123, 703e734.
Whattam, S.A., 2019. Primitive magmas in the early Central American Volcanic Arc
system generated by plume-induced subduction initiation. Front. Earth Sci. 6,
114. https://doi.org/10.3389/feart.2018.00114.
Whattam, S.A., Malpas, J., Ali, J.R., Smith, I.E.M., 2008. New SW Pacific tectonic
model: cyclical intraoceanic magmatic arc construction and near coeval
emplacement along the AustraliaePacific margin in the Cenozoic. Geochem.
Geophys. Geosyst. 9 https://doi.org/10.1029/2007GC001710.
Whattam, S.A., Malpas, J., Smith, I.E.M., Ali, J.R., 2006. Link between SSZ ophiolite
formation, emplacement and arc inception, Northland New Zealand: UePb
SHRIMP constraints; Cenozoic SW Pacific tectonic implications. Earth Planet.
Sci. Lett. 250, 606e632.
€rner, G., Harmon, R.S., Hartmann, G., Simon, K., 2005. Igneous geology and
Wo
geochemistry of the upper Río Chagres basin. In: Singh, V.P., Harmon, R.S. (Eds.),
The Río Chagres, Panama: A Multidisciplinary Profile of a Tropical Watershed.
Springer, Dordrecht, Netherlands, pp. 65e81.
Whattam, S.A., Montes, C., McFadden, R.R., Cardona, A., Ramirez, D., Valencia, V.,
2012. Age and origin of earliest adakitic-like magmatism in Panama: implica-
tions for the tectonic evolution of the Panamanian magmatic arc system. Lithos
142e143, 226e244.
Whattam, S.A., Stern, R.J., 2011. The ‘subduction initiation rule’: a key for linking
ophiolites, intra-oceanic forearcs, and subduction initiation. Contrib. Mineral.
Petrol. 62, 1031e1045. https://doi.org/10.1007/s00410-011-0638-z.
Whattam, S.A., Stern, R.J., 2015a. Late Cretaceous plume-induced subduction initi-
ation along the southern margin of the Caribbean and NW South America: the
first documented example with implications for the onset of plate tectonics.
Gondwana Res. 27, 38e63.
Whattam, S.A., Stern, R.J., 2015b. Arc magmatic evolution and the construction of
continental crust at the Central American Volcanic Arc system. Int. Geol. Rev.
https://doi.org/10.1080/00206814.2015.1103668.
Wilson, J.T., 1966. Did the Atlantic close and re-open? Nature 211, 676e681.
Winchester, J.A., Floyd, P.A., 1977. Geochemical discrimination of different magma
series and their differentiation products using immobile elements. Chem. Geol.
20, 325e343.
Wo€rner, G., Harmon, R.S., Wegner, W., 2009. Geochemical evolution of igneous
rocks and changing magma sources during the evolution and closure of the
Central American landbridge. In: Kay, S.W., Ramos, V.A., Dickinson, W. (Eds.),
Backbone of the Americas: Shallow Subduction, Plateau Uplift and Ridge and
Terrane Collision, vol. 204. Geological Society of America Memoir, pp. 183e196.
Zhou, X., Li, Z.-H., Gerya, T.V., Stern, R.J., Xu, Z., Zhang, J., 2018. Subduction initiation
dynamics along a transform fault control trench curvature and ophiolite ages.
Geology 46, 607e610. https://doi.org/10.1130/G40154.1.